Up-regulation of vitamin D receptor (VDR) expression has been shown in several tumors and is thought to represent an important endogenous response to tumor progression. The authors aimed to verify the expression of VDR and its clinical significance in histologically proven cholangiocarcinoma (CCA).
The antiproliferative activity of vitamin D3 on CCA cell lines was explored. The immunohistochemistry of 111 paraffin-embedded CCA tissues showed that VDR expression gradually increased during CCA development. Normal bile duct epithelium rarely expresses VDR, whereas more than 74% of CCA tissues showed positive VDR staining, of which 40% were high. Approximately 80%–90%of CCA patients with papillary and well differentiated adenocarcinomas had positive VDR expression in tumor tissues, whereas 39% positive VDR expression was found in those with poorly differentiated CCAs (P < .001).
Expression of VDR was shown to be compatible with an overall favorable prognosis for CCA. Treatment with 1,25(OH)2D3, an active metabolite of vitamin D3, in the CCA cell lines with high expression of VDR significantly reduced cell proliferation in a dose-dependent manner. The effect was not demonstrated in the CCA cell lines that had lower VDR expression.
The active metabolite of vitamin D3—1alpha, 25-dihydroxyvitamin D3 (1,25(OH)2D3)—helps maintain mineral homeostasis and normal skeletal architecture. Apart from these calcium-related actions, 1,25(OH)2D3 is also recognized as a regulator of growth and differentiation of many tissues such as skin, muscle, and heart.1 The actions of 1,25(OH)2D3 are mediated through vitamin D receptors (VDRs), which belong to the steroid/thyroid hormone nuclear receptor super family. 1,25(OH)2D3 and its receptors form a complex that binds to vitamin D-responsive elements, and it either positively or negatively affects transcription.2 The VDR ligand-binding domain acts as a molecular switch, because its interactions with corepressor or coactivator proteins are the central molecular events that trigger nuclear vitamin D3 signaling.
An association among 1,25(OH)2D3, VDR, and cancer has been recognized since epidemiological studies indicated an inverse relation between levels of vitamin D3 and risk of many types of cancer.3 In addition, 1,25(OH)2D3 has been found to regulate differentiation and growth of several cultured human cell lines.4, 5 Moreover, the protective effects of 1,25(OH)2D3 against cancers are evident in neoplasms of the colon, breast, prostate, and other sites.6, 7
The physiological and pharmacological actions of 1,25(OH)2D3 in many pathologic conditions, along with the detection of high VDR expression in target cells, have signaled potential therapeutic applications of VDR ligands for osteoporosis, dermatological indications, inflammation, autoimmune diseases, and cancers (ie, prostate, colon, breast, myelodysplasia, leukemia, head and neck squamous cell carcinoma, and basal cell carcinoma). Consequently, many 1,25(OH)2D3 analogs with noncalcific actions have been developed, with the goal of improving the biological profile of the natural hormone for its therapeutic application for antiproliferative, prodifferentiative, and immunomodulatory effects.8, 9
Cholangiocarcinoma (CCA) is a malignant neoplasm arising from the biliary epithelium, either within the intrahepatic or extrahepatic biliary tract. Even though CCA is rare worldwide, the incidence and mortality rates of CCA in the past 3 decades have grown in the USA, the UK, Japan, and Australia.10, 11 Because its prognosis is poor and effective therapy is lacking, CCA is usually a fatal cancer. Therefore, novel treatment strategies directed against this malignancy constitute an urgent task.
Molecular alterations associated with CCA have been made by using different high-throughput techniques.12, 13 Exploiting select molecular targets that are aberrantly expressed during carcinogenesis and metastasis of CCA is 1 approach being used to develop specific chemopreventive and therapeutic strategies. Accordingly, we first applied a comparative genomics approach by using the serial analysis of gene expression (SAGE) database (http://cgap.nci.nih.gov/SAGE). A comparison of SAGE data of tissues obtained from metastatic CCA and normal liver indicated that several genes are up-regulated in CCA tissues. We focused on the vitamin D receptor (VDR) gene (SAGE tag: GAGAAACCCT) because 1,25(OH)2D3 has recently been used in chemoprevention and therapeutics for human tumors.8, 9
Thus, the expression of VDR in tumor tissues was investigated from 111 patients with histologically proven CCAs. The association of VDR expression and clinical findings from patients, and the antiproliferative activity of 1,25(OH)2D3 on CCA cell lines, were determined. Here, we show for the first time that VDRs are aberrantly expressed during development of CCA and that the expression of VDRs is a good prognostic marker of CCA. Supplementation of 1,25(OH)2D3 significantly reduced proliferation of CCA cells. In addition, the antiproliferative effects of 1,25(OH)2D3 demonstrated in CCA cell lines were mediated by VDR. These findings suggest that supplementation of 1,25(OH)2D3 or its analogs may be a potential strategy for long-term control of tumor development and progression in CCA patients.
MATERIALS AND METHODS
Subjects and Tissues
Paraffin-embedded liver tissues of patients with histologically proven CCA who underwent liver resection (n = 111) were obtained from the specimen bank of the Liver Fluke and Cholangiocarcinoma Research Center, Faculty of Medicine, Khon Kaen University, Thailand. Informed consent was obtained from each subject, and the Khon Kaen University Ethics Committee for Human Research approved the research protocol (HE 471214). Normal and hyperplastic/dysplastic biliary epithelia were examined from the noncancerous portion of liver tissues obtained from CCA patients. Tumor location (n = 99), tumor size (n = 82), pTNM stage,14 and histological grading (n = 111) were evaluated by reviewing medical charts and pathology records.
Cell Lines and Proliferation Assay
Four CCA cell lines, M139, M213, OCA17, and KKU100, were established as described.15 Cells were cultured in Ham F12 supplemented with 10% heat-inactivated fetal calf serum, 1% L-glutamine, and 1% penicillin-streptomycin at 37°C and 5% CO2.
CCA cells (3 × 103 cells/well) were plated in a 96-well culture plate for 24 hours. Media with various concentrations of 1,25(OH)2D3 (Sigma-Aldrich, St. Louis, Mo) were added to a final volume of 100 μL per well. Plates were incubated further for 72 hours, and cell numbers were determined by using sulforhodamine B assay (SRB, Sigma-Aldrich, St. Louis, Mo). Briefly, the culture medium was removed, 10% cold trichloroacetic acid was added for 30 minutes at room temperature, and subsequently washed 5 times with deionized water. The plate was air dried, and 0.4% SRB in 1% acetic acid was added for 30 minutes. Unbound dye was washed out 5 times with 1% acetic acid. After air drying, SRB dye within cells was solubilized with 200 μL of 10 mM unbuffered Tris-base solution. The plate was shaken for at least 10 minutes, and absorbance was measured at 540 nm by using a microplate reader (Tecan Austria GmbH, Salzburg, Austria).
Semiquantitative Polymerase Chain Reaction (PCR)
The total RNA of the CCA cell line was extracted by using RNeasy Kit (Qiagen, Tokyo, Japan), and generated to cDNA by Ready to Go reverse transcription polymerase chain reaction (RT-PCR) Beads (Amersham Pharmacia Biotech, Buckinghamshire, England), according to the manufacturer's instructions. Single tube, semiquantitative, RT-PCR analysis was used to determine mRNA expression levels of VDR and glyceraldehyde-3-phosphatedehydrogenase (GAPDH) (The latter was used as an internal standard.) The number of PCR cycles was determined to yield a linear range of PCR amplification products for VDR and GAPDH. The following specific PCR primers for VDR and GAPDH mRNAs were used, VDR forward primer, 5′-TGT CCT GGA CCT GTG GCA AC-3′; VDR reverse primer, 5′-AGG ACG ATC TGG GGA GAC GA-3′ and GAPDH forward primer 5′-AAT GCC TCC TGC ACC ACC-3′; reverse primer 5′-CTG TTG AAG TCA GAG GAG AC-3′. The sizes of the generated PCR products were 182 base pairs (bp) for the VDRs and 416 bp for the GAPDH. The thermal cycle conditions for VDR and GAPDH included 1) denaturation at 94°C for 2 minutes; 2) 35 cycles of denaturation at 94°C for 30 seconds; 3) annealing at 65°C for 30 seconds; 4) extension at 68°C for 1 minute; and, 5) a final extension at 68°C for 10 minutes (MBS satellite 0.2G Thermal cycler; Thermo Electron, Milford, Mass). The PCR product (5 μL) was separated on 2.0% agarose gel. The intensity of each band was quantified by using GelPro 32 (Media Cybernetics, Bethesda, Md). Water was used as a negative control instead of a cDNA product in order to reveal any false-positive reactions.
VDR protein was detected on formalin-fixed, paraffin-embedded sections by using the standard immunohistochemistry technique. Briefly, deparaffinized tissue sections were blocked with 3% H2O2 and nonimmune horse serum for 30 minutes. Afterward, slides were heated in a boiling pressure in 10 mM citrate buffer, pH 6.0, for 4 minutes. The slides were allowed to react at room temperature overnight with 1:400 rat anti-VDR monoclonal antibody (Chemicon; Temecula, Calif). The slides were then incubated with 1:100 goat antirat IgG (H+L) (Zymed Laboratories; San Francisco, Calif). The peroxidase activity was observed by using diaminobenzidine tetrahydroxychloride solution (DAB; Dako; Glostrup, Denmark) as the substrate. The sections were counterstained with hematoxylin. The positive staining was eliminated when phosphate-buffered saline was applied instead of the primary antibody. The staining frequency of VDR was semiquantitatively scored on the basis of the percentage of positive cells as: 0 = negative; 1+ = 1% to 25%; 2+ = 26–50%; and 3+>50%. In some analyses, the VDR expressions of 1+, 2+, and 3+ were summed as for cumulative positive VDR expression.
The analyses were performed by using SPSS 14.0 for Windows Evaluation software (SPSS Inc., Chicago, Ill). Fisher exact test was used to evaluate the association between VDR expression and clinical factors of CCA patients. The cumulative survival after tumor removal was calculated according to the Kaplan-Meier method, with a log-rank test. A multivariate survival analysis was performed on parameters found to be significant in univariate analysis by the Cox proportional hazard regression model. P < .05 was considered statistically significant.
Of the 111 CCA patients examined, 78.37% were men and the ratio of men to women was 3.6 to 1. The median age of the CCA patients was 55 years (range, 29 to 73 years). In our series, peripheral CCA (78.68%) was the major type, and most of the tumors (87.70%) were at an advanced stage (ie, IVA or IVB). Survival of each patient after surgery was recorded until April 27, 2006. Ninety percent of the patients died by the end of the follow-up period.
The operative procedures for all patients were intended for cure; however, the pathology record showed 58.5% (65 of 111) of cases were margin free. In this series, 18.9% (21 of 111) of patients received complete chemotherapy after surgery, and none of the patients received postoperative radiotherapy.
Expression of VDR in Bile Duct Epithelia and CCA Tissue
VDR was rarely expressed in normal bile duct epithelia, and only 20% of normal bile ducts showed positive staining with low expression (Fig. 1a). In contrast, VDR was expressed more frequently in precancerous (hyperplasia and dysplasia) biliary epithelial cells and CCA tissues than in normal bile duct epithelia (P < .001) (Table 1). The frequency of positive staining of VDR was gradually increased in hyperplasia to dysplasia and CCA (Fig. 1b-d) (Table 1). More than 74% of CCA tissues showed positive VDR staining, 40% of which had highly expressed VDR. In all cases, VDR was mostly detected in cytoplasm.
Table 1. Expression of VDR in Bile Duct Epithelium and CCA Tissues
Bile duct epithelium
Correlation Between VDR Expression and Clinicopathologic Features
By using a univariate analysis, the association of VDR expression in CCA tissues with patients' clinicopathology was determined. The expression of tissue VDR had no association with age, sex, tumor type, tumor stage, tumor size, or metastasis; however, it was significantly correlated with histological grading (Fig. 2). The proportion of patients with positive VDR was significantly higher in those with papillary (15 of 18, P < .001), well differentiated (47 of 51, P < .001), and moderately differentiated (10 of 14, 71%) (P < .05) adenocarcinomas than those with a poorly differentiated type (11 of 28, 39%).
Correlation Between VDR Expression and Cumulative Survival Rates
Cumulative survival was compared between CCA patients; those with and without positive VDR expression. Patients with survival less than 30 days were excluded from analysis and were considered perioperative deaths.
The log-rank analysis indicated that CCA patients with positive tumor VDR expression had significantly better survival than those with negative VDR expression (P = .003, Fig. 3). Respective mean survival times in days were 330 ± 48 (95% CI = 235–425 days) and 162 ± 25 (95% CI = 112–211 days).
A multivariate progression analysis, based on the Cox proportional hazard model, was performed to analyze the independent value of each parameter predicting overall survival (Table 2). Factors that may affect survival were included in the model. In this analysis, VDR expression in tumor tissue (P = .03), age of CCA patients (P = .039), and complete postoperative chemotherapy (P = .001) proved to be independent prognostic factors for longer overall survival in CCA. CCA patients who had no VDR expression had a 2-fold higher risk of death than patients whose tumors expressed VDR.
Table 2. Multivariate Analysis by a Cox Proportional Hazards Regression Model
Antiproliferative Activity of 1,25(OH)2D3 in CCA Cell Lines
In the current study, 4 human CCA cell lines—namely M139, M213, KKU100, and OCA17—were selected to assess the role of VDR in the antiproliferative activity of 1,25(OH)2D3 in CCA. VDR expression levels in CCA cell lines were determined by semiquantitative PCR, using GAPDH as the internal standard. VDR expression of cell line M213 was significantly higher than those of cell lines M139, KKU100, and OCA17 (P < 0.001) (Fig. 4).
All cell lines were treated by different concentrations of 1,25(OH)2D3 for 3 days. 1,25(OH)2D3 significantly decreased proliferation of M139 and M213 (P < .05) in a dose dependent manner (Fig. 5); by comparison, the CCA cell lines KKU100 and OCA17 did not respond.
The aberrant expression of VDR during development of CCA was addressed in this study through an immunohistochemical examination of VDR expression in 111 CCA tissues. A higher expression of VDR was found in bile duct epithelium with hyperplasia/dysplasia and CCA tissues: More than 74% of primary CCA tissues overexpressed VDR. Up-regulation of VDR has been reported also in primary tumor tissues of breast,16 colon,17 and pancreatic cancers.18 High expression of VDR in preneoplastic lesions—established tumors and lung metastases of MMTV-neu transgenic mouse model of breast cancer—also have been demonstrated.19 Aberrant expression of VDR in hyperplastic and/or dysplastic biliary cells, as reported in the present study, may be part of early molecular events that were altered during malignant formation of CCA. However, this observation is only correlative; investigations into the molecular mechanism(s), which dysregulates normal VDR signaling in CCA development, are needed. This may generate opportunities for therapeutic intervention, achieved through disruption of CCA tumorigenesis.
In the present study, there is no significant correlation of VDR expression with tumor size, tumor stage, or metastasis. This suggests that VDR expression may not associate with tumor progression of CCA. However, the expression of VDR in CCA tissues is significantly correlated with histological type of CCA. More than 80% of CCA patients with papillary type and 90% of well differentiated type CCA had positive VDR expression (Fig. 2). Both histological types are known to possess good prognosis compared with poorly differentiated types.20, 21 This observation implies that high expression of VDR may correlate with good prognosis CCA. This postulation may be supported by the fact that 1,25(OH)2D3, and its potent analogs, can induce growth arrest and differentiation of some cancer cells.22, 23 In the present study, expression of VDR and postoperative chemotherapy were found to be independent prognostic variables for better survival of CCA patients, regardless of histological type, tumor type, tumor stage, and metastasis, (Table 2). Patients who were positive for tumor VDR expression had significantly longer survival than those who were not. A similar observation has also been reported for patients with colon cancer24 and esophageal cancers.25
The antiproliferative activities of 1,25(OH)2D3 were determined in 4 CCA cell lines with high (M139, M213) and low (OCA17, KKU100) expressions of VDR. In a dose-dependent fashion, 1,25(OH)2D3 significantly suppressed growth of CCA cell lines M139 and M213 although not in cell lines OCA17 and KKU100. Cells with higher numbers of VDR tended to be more sensitive to the effect of 1,25(OH)2D3 on growth, in agreement with other reports.18, 26, 27 The positive correlation between VDR expression levels and the antiproliferative activity of 1,25(OH)2D3, as demonstrated in the current study, also has been reported on several other cancer cell lines.26 Transfection of MCF-7 breast cancer cells with antisense VDR significantly reduced antiproliferative sensitivity28 to 1,25(OH)2D3. In contrast,29 transfection of VDR expression vector to HBL100, a vitamin D resistant human breast epithelial cell line, enhanced sensitivity of cells to the growth inhibitory effect of 1,25(OH)2D3.
Vitamin D3 is a seco-steroid hormone that regulates calcium homeostasis within the body. It is also recognized as a regulator of growth and differentiation of many tissues. Vitamin D3 exerts its activity through both genomic and nongenomic pathways. The genomic actions of 1,25(OH)2D3 are modulated through the VDR. Most dividing cell types, normal and malignant, can express VDR and respond to 1,25(OH)2D3. Significant antineoplastic activity of 1,25(OH)2D3 has been shown in preclinical models; of which different mechanisms have been proposed according to tumor models and experimental conditions. These include inhibition of proliferation associated with cell cycle arrest,23, 30, 31 differentiation,32, 33 reduction invasiveness and angiogenesis,34, 35 and stimulation of apoptosis.36 In the present study, cells treated with 1,25(OH)2D3 induced apoptosis as observed in CCA cell line M213.
In addition to genomic effects, vitamin D also regulates a number of cytoplasmic signaling pathways, such as protein kinase; ras and mitogen-activated protein kinase (MAPK).37 These actions often result in rapid changes in intracellular calcium and activation or deactivation of a number of proteins, which finally affect cellular growth, differentiation, and apoptosis. Several studies, however, suggest that VDR is required for expression of gene products involved in nongenomic response. The exact role of VDR in this action remains to be determined.
1,25(OH)2D3 (calcitriol) and multiple analogs with reduced calcific properties are under investigation in numerous cultured cell types26, 38, 39 and in several in vivo animal models.40 Calcitriol, the principal biologically active ligand of VDR has been shown to inhibit cancer cell proliferation in both in vitro and in vivo models of a wide range of neoplasms. Efficacy of calcitriol and its analogs on cancer prevention and therapy has been explored in prostate cancer,41, 42 colon cancer,43 and breast cancer.44
In the present study, the genomic action of vitamin D3 through VDR is emphasized. The association of VDR expression in CCA tissue indicating a good prognosis, together with the reduction of cell proliferation by 1,25(OH)2D3 treatment, suggest the possibility of using vitamin D as an adjuvant therapy for CCA patients who have high expression of VDR. However, investigation of a mechanism by which VDR and its ligand mediate these processes is needed to provide the basis for the potential use of this hormone and its derivatives in the prevention and treatment of CCA.
We thank Mr. Bryan Roderick Hamman for assistance with the English-language presentation of this article.