Background
Prostate cancer is the most common malignant tumor in men in many industrialized nations and is the second highest cause of cancer mortality [
1]. As prostate cancer relies on androgens for its maintenance and progression, targeting androgens offers a therapeutic opportunity to halt or delay the progression of prostate cancer. Androgen-deprivation therapies, including castration (orchiectomy, the surgical removal of the testicles) and pharmacological control, are the mainstay for the management of advanced prostate cancer, reducing symptoms in approximately 70–80% of patients. Regardless, most tumors relapse within two years to an incurable hormone-independent state, in which tumor cells are scarcely responsive to even high concentrations of chemotherapeutic agents or radiotherapy [
2]. The regulatory mechanisms that cause this transition remain largely unknown, and no effective therapy for androgen-independent prostate cancer (AIPC) has been developed to date.
DNA methylation is an important regulator of gene transcription and plays an important role in the process of tumorigenesis and the development of prostate cancer [
3,
4]. 5-Aza-2’-deoxycytidine (5-Aza-CdR), a nucleoside analog inhibitor of DNA methyltransferase (DNMT), has been used to reverse methylation and reactivate the expression of silenced genes [
5]. 5-Aza-CdR was able to suppress the growth of various tumors in vitro, animal models, and clinical trials including prostate cancer [
6‐
9], hematopoietic malignancies [
10,
11], and lung carcinoma [
12]. In vitro studies investigating the effects of 5-aza-CdR in prostate cancer cell lines demonstrated that a low-dose 5-aza-CdR treatment regimen given daily completely inhibited cell proliferation and induced cell death in LNCaP and PC3 cell lines [
13]. Moreover, several studies have been conducted to examine the synergistic effects of 5-Aza-CdR and chemotherapeutic agents against tumor cells. A combination of 5-Aza-CdR and cisplatin showed synergy in triggering the apoptotic death of PC3 cells [
14], and the combined treatment of castration and 5-Aza-CdR was more effective than either treatment alone according to tissue histology, significantly prolonging survival in a transgenic adenocarcinoma mouse model of prostate cancer [
9]. Nevertheless, it remains unknown whether 5-Aza-CdR can delay the progression of castration-resistant prostate cancer, and little is known about the genes silenced in progressed prostate cancer cells that are reactivate by 5-Aza-CdR.
MicroRNAs (miRNAs) are a class of short (between 19 and 25 nucleotides) noncoding RNAs that negatively regulate gene expression via complementary binding to target messenger RNA (mRNA) [
15], thereby impairing the translation of the mRNA or marking it for early degradation. Dysregulated miRNA expression plays an important role in the process of tumorigenesis and development of androgen-independent tumors after castration in prostate cancer patients [
16]. A recent study reported that the excessive expression of miR-146a was exclusively found in prostate cancer LNCaP and PC3 cell lines [
17]. Transfection of miR-146a into a cell line was found to significantly suppress the expression of ROCK1, consequently markedly reducing cell proliferation, invasion, and metastasis to human bone marrow endothelial cell monolayers [
17,
18], indicating that miR-146a may function as a tumor-suppressor gene in the transformation of AIPC and metastasis in prostate cancer.
In the present study, we found that 5-Aza-CdR could inhibit the cell viability of either androgen-dependent LNCaP cells or androgen-independent PC3 cells through the up-regulation of miR-146a expression. To evaluate the effect of 5-Aza-CdR on the progression of AIPC in vivo, we established subcutaneous LNCaP xenografts in castrated male mice. As expected, 5-Aza-CdR delayed the progression of castration-resistant tumors. The methylation levels of the global genome and miR-146a promoter were decreased after 5-Aza-CdR treatment, which resulted in the increased expression of miR-146a, suggesting that castration in combination with 5-Aza-CdR may be a new therapeutic option for both androgen-dependent and -independent prostate cancer.
Methods
Cell culture and treatment
LNCaP and PC3 human prostate carcinoma cells (American Type Culture Collection, ATCC) were maintained in RPMI 1640 supplemented with 10% fetal bovine serum (FBS), 100 U/ml penicillin, and 100 pg/ml streptomycin (Gibco, NY, USA). 5-Aza-CdR (Sigma-Aldrich, St. Louis, Mo, USA) was dissolved in dimethylsulfoxide (DMSO). Fresh medium containing 5-Aza-CdR was replaced every 24 h. MiR-146a inhibitors (single-stranded chemically modified oligonucleotides; Life Technologies Corporation, Shanghai, China) were used for the inhibition of miR-146a expression in LNCaP and PC3 cells. Negative control inhibitors (ctrl inhibitors) were transfected as matched controls. Cells were transfected with RNAs using INTERFERin (Polyplus-Transfection SA, Illkirch, France) according to the manufacturer’s instructions.
Cell viability assays
LNCaP and PC3 cells were seeded in 96-well culture plates with 100 μl of growth medium. Following 5-Aza-CdR or miR-146a inhibitor treatment, MTS cell viability assays were performed according to the manufacturer’s instructions.
LNCaP xenograft model studies
Male BALB/c nude mice at 4–6 weeks age (18–20 g) were obtained from the Institute of Zoological Sciences, Chinese Academy of Medical Sciences in Beijing. For the tumor growth studies, 5 × 106 LNCaP cells in 0.1 ml suspension were mixed with 0.1 ml of Matrigel (Collaborative Research Inc., Bedford, MA, USA) and inoculated into the right dorsal flanks of the mice. When the tumor volumes reached approximately 250 mm3, the animals were randomly divided into four groups (n = 12). One was administered thrice weekly on consecutive days intraperitoneal (i.p.) injections of 0.25 mg/kg 5-Aza-CdR dissolved in PBS. Two groups underwent bilateral orchiectomy (castration) under metofane (Pitman-Moore, Washington Crossing, NJ) anesthesia. Three days after surgery, these two groups were injected with either 5-Aza-CdR or PBS thrice weekly i.p. The control mice were testis-intact animals, with injections of PBS. The tumor volume was measured with calipers twice weekly for length and width; the volume was then calculated by the formula: [(length × width2)/2]. The animals of each group were euthanized on the 14th and 35th day post-treatment (n = 6 per group at either time point), respectively. All animal experiments were approved by the Committee of Use and Care of Animals, Chinese Academy of Medical Sciences.
Measurement of miRNA and mRNA expression
The small and total RNA fractions were isolated isolated from cells and tumor tissue using the miRVana miRNA Isolation Kit (Ambion Inc, Austin, TX, USA) according to the manufacturer’s instructions. MiRNA was reverse-transcribed using the TaqMan microRNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA, USA) in a reaction mixture containing an miR-specific stem-loop reverse transcription (RT) primer. The Taqman miRNA assay system (Applied Biosystems, Foster City, CA) was used according to the manufacturer’s instructions to quantitatively detect the expression of mature miRNAs was performed using the TaqMan miRNA Assay Kit (Applied Biosystems) containing TaqMan primers in a universal PCR master mix. CDNA was synthesised by reversetranscription using ReverTra Ace (Toyobo, Osaka, Japan). The SYBR Green PCR Master Mix (Toyobo, Osaka, Japan) was used to analyze mRNA expression. Quantification of miRNAs and mRNA by qRT-PCR was performed using an ABI 7300HT thermocycler (Applied Biosystems) at 95°C for 10 min, followed by 40 cycles of 95°C for 15 s and 60°C for 1 min. Reactions were performed in triplicate with human U6 or GAPDH as an internal control. The primer sequences for real-time quantitative PCR (qRT-PCR) are shown in Table
1. Fluorescent signals were normalized to an internal reference, and the threshold cycle was set within the exponential phase of PCR. The tumor tissue from the control mice of the first analysis was used to the calibrator samples. The relative gene expression was calculated by comparing cycles for each target PCR. Cycle threshold values were converted to relative gene expression levels using the 2
-ΔΔCT method.
Table 1
Primers for PCR and BSP
DNMT1 | F: TACCTGGACGACCCTGACCTC |
R: CGTTGGCATCAAAGATGGACA |
DNMT3a | F: TATTGATGAGCGCACAAGAGAGC |
R: GGGTGTTCCAGGGTAACATTGAG |
DNMT3b | F: GGCAAGTTCTCCGAGGTCTCTG |
R: TGGTACATGGCTTTTCGATAGGA |
GAPDH | F: CGACCACTTTGTCAAGCTCA |
R: AGGGGTCTACATGGCAACTG |
LINE1 | F: CCGTAAGGGGTTAGGGAGTTTTT |
R: RTAAAACCCTCCRAACCAAATATAAA |
HmiR-146a promoter | F: AGGGAGTTTTTTGTTTGATTTTTTTT |
R: CTATCCACCCTTTAACATACCTTCC |
Western blot analysis
Cells were lysed for 30 min at 4°C in PBS with 1% NP-40 and protease inhibitor cocktail tablets (Roche, Mannheim, Germany). Protein concentration was assayed by using bicinchoninic acid (Pierce, Rockford, IL). Total proteins were separated by electrophoresis on 12% SDS-polyacrylamide gels and 4% polyacrylamide gels and transferred to nitrocellulose. Immunodetection of caspase 3 and GAPDH was carried out by using anti-caspase 3 antibody at a dilution of 1:400 and anti-GAPDH (Cell Signaling Technology, Inc., Beverly, MA, USA). Specific proteins were visualized with enhanced chemiluminescence reagents (Hyperfilm ECL; Amersham Biosciences, Buckinghamshire, England) followed by exposure to X-ray film for 1 to 3 min.
ELISA assay
Blood samples were collected by tail vein incision, and the serum was stored at −20°C until assayed. The free prostate-specific antigen (FPSA) in mouse serum was measured using ELISA kits (R&D Systems, USA) according to the manufacturer’s protocol.
COBRA assay
The bisulfite modification of genomic DNA was performed using the CpGenome Turbo Bisulfite Modification Kit (Billerica, MA, USA), and the methylation analysis of LINE-1 repetitive elements was performed initially using the COBRA assay. The primer sequences that correspond to the nucleotides in the regulatory region of the LINE-1sequence are shown in Table
1. The PCR reactions consisted of 35 cycles of 95°C for 1 min, 53°C for 1 min, and 72°C for 1 min. The PCR products were subsequently digested with
TaqI (MBI Fermentas) and
TasI (MBI Fermentas) in TE buffer3 (Biolab) at 65°C overnight and then separated by electrophoresis on polyacrylamide gels. The gel was stained by silver staining. The methylated bands of LINE-1 (
TaqI positive) yielded two 80-bp DNA fragments, whereas the unmethylated amplicons (
TasI positive) yielded 63- and 97-bp fragments. The methylation levels were calculated as the intensity of methylated bands divided by the sum of the methylated and unmethylated bands. Each COBRA assay was performed two to four times.
Quantification of the methylation levels of the miR-146a promoter
The miR-146a promoter regions essential for basal transcriptional activity occur between nt −442 and nt +5 and include 11 CpG sites. The 447-bp region was amplified by bisulfite specific PCR (BSP) using primers designed with the MethPrimer software (Table
1). The PCR reaction contained modified genomic DNA and was overlaid with mineral oil to form a vapor barrier. The cycling conditions consisted of an initial denaturation at 96 ºC for 5 min after which 0.5 μl reverse primer (10 μM) was added to the PCR mixture, followed by 2 cycles of 96°C for 1 min, 58°C for 2 min, and 72°C for 2 min. When the temperature again rose to 96°C, which was at the beginning of the next phase of 8 cycles (96°C for 1 min, 58°C for 2 min, and 72°C for 2 min), 0.5 μl forward primer was added to the mixture. Then, 30 cycles of 96°C for 30 s, 55°C for 45 s, and 72°C for 45 s were performed, followed by a final extension of 72 ºC for 7 min. The PCR products were sequenced using a DNA sequencer (ABI PRISM 3730, Foster City, CA, USA), and the methylation levels were measured according to our previously reported method [
19].
Statistical analysis
All the data were obtained from at least three independent experiments. The data were expressed as the mean values ± SD and compared between two groups using Student’s t-test. The statistical significance was defined as a P value less than 0.05.
Discussion and conclusions
After prostate cancer patients undergo androgen deprivation as the first-line initial treatment, most patients develop to a more aggressive and androgen-independent status, with a median survival from 2 to 3 years. However, as the current clinical treatments for androgen-independent disease are not ideal, a therapy that either prevents or delays the development of AIPC is needed for these patients. In the present study, we found that 5-Aza-CdR could delay the emergence of lethal androgen-independent tumors in response to castration. Moreover, 5-Aza-CdR could augment miR-146a expression by decreasing the methylation levels of the miR-146a promoter in an LNCaP xenograft model of castrated mice, a result that was associated with the inhibition of prostate cancer progression. This is the first study to reveal potential molecular alterations by DNA methylation inhibitors as an effective therapy for preventing the relapse of castration-resistant prostate cancer.
The conversion to an androgen-independent phenotype is a complex process involving multiple molecular mechanisms. Prostate cancer cells can survive at low levels of androgens in castrated patients via androgen receptor (AR) mutation or amplification [
24], or by increasing 5a-reductase activity to convert testosterone to dihydrotestosterone [
25]. In addition, epigenetic alterations are the most common genome changes in prostate cancer cells and are associated with defects in gene function, contributing to carcinogenesis and helping the cancer cells survive and grow under androgen-independent conditions. Many specific genes are hypermethylated and inactivated during prostate cancer progression, including
APC, MDR1, GPX3, and
p16[
26], and alterations in DNA methylation are now being used as molecular biomarkers for prostate cancer detection, diagnosis, and prognosis. 5-Aza-CdR, a DNMT inhibitor, can remove methyl residues from silenced genes, resulting in re-expression, and has been approved for use against hematopoietic malignancies. This inhibitor has also shown clinical efficacy in the treatment of metastatic lung carcinoma, acute lymphoid leukemia, chronic myeloid leukemia, and head and neck cancer. Furthermore, 5-Aza-CdR has been evaluated in hormone-independent metastatic prostate cancer patients in a phase II trial by Thibault
et al. [
27], and several studies have indicated the effect of 5-Aza-CdR on prostate cancer in animal models. Studies concerning combination chemotherapy of 5-Aza-CdR and chemotherapeutic agents have been performed, with the results suggesting the synergistic growth suppression of 5-Aza-CdR plus PTX in all PC cell lines [
28]. In the present study, 5-Aza-CdR induced apoptosis of the androgen dependent (LNCaP) and androgen-independent (PC3) prostate cancer cells. The excessive expression of miR-146a and the demethylation of its promoter were exclusively found in both cell lines treated with 5-Aza-CdR, which was involved apoptosis of prostate carcinoma. Then we paid attention to if 5-Aza-CdR could delay the progression of prostate cancer to androgen independence in vivo.
Christoph
et al. ever using a transgenic adenocarcinoma mouse model of prostate cancer demonstrated that the combined treatment of 5-Aza-CdR plus castration significantly prolonged survival and that 5-Aza-CdR appeared to delay the onset of androgen-independent disease [
9]; however, that animal model did not mimic the course of the clinical disease. Nonetheless, the molecular mechanism by which 5-Aza-CdR inhibits the progression of prostate cancer remains unclear. Of the available human prostate cancer cell lines, only the LNCaP cell line is androgen responsive, PSA secreting, and immortalized in vitro. Thalman
et al. demonstrated that the LNCaP progression model, unlike other human prostate cancer models, shares remarkable similarities with human prostate cancer and can transform a subpopulation to AI clones in response to acute or chronic androgen ablation. Based on previous studies, we utilized the LNCaP xenograft model to evaluate the effect of 5-Aza-CdR on the growth suppression of prostate cancer in castrated male mice. We observed that, within the initial 17 days, the xenograft tumor volumes were decreased in response to castration compared to the control, though the tumor volumes began to increase in the next stage, illustrating that the LNCaP xenograft model can mimic the course of the clinical states of the transition from initial androgen dependency to androgen independency, which is consistent with other reports [
29,
30]. 5-Aza-CdR treatment significantly suppressed tumor progression and delayed the regeneration of prostate cancer cells in the later stage compared to the castration treatment alone. Correspondingly, FPSA concentrations, an important clinical marker of prostate cancer progression, were also significantly down-regulated at the second stage in the combination treatment group compared to that in either the 5-Aza-CdR or castration alone groups. Although our study showed an effect of 5-Aza-CdR on the delay of the onset of androgen-independent disease in the mouse model, further research is necessary to evaluate whether 5-Aza-CdR can prolong survival and how long 5-Aza-CdR can delay the onset of androgen-independent disease in the castrated mouse model.
Additionally, we found that the expression of miR-146a was negatively correlated with tumor volumes, which significantly enhanced in the tumor xenografts of 5-Aza-CdR-treated mice and the androgen-dependent stage but not the androgen-independent stage of castrated mice, compared to control mice. In particular, the expression level of miR-146a was highest in the combined treatment (castration and 5-Aza-CdR), either androgen-dependent or androgen-independent stage. We further found the change trend of NF-κB activity, which is one of critical transcription factors for miR-146a expression, was consistent with miR-146a expression in tumor tissues of different groups (data not shown). Unfortunately, we could not reveal the potential mechanism involved in the regulation of NF-κB activity by 5-Aza-CdR or castration in the xenograft tumors of mice.
Since DNMT1, DNMT3A and DNMT3B are known to be involved in de novo methylation and the maintenance of methylation patterns of genes, we investigated the expression levels of all three members of the DNMT family. Our data clearly indicates that the DNMTs mRNA levels were all decreased after castration at the initial stage, yet were little changed at the later stage in different groups compared to the control in castrated mice. There was no a correlation between the levels of DNMT mRNA expression and the extent of global genomic DNA methylation, as measured by LINE-1 repetitive sequences. We deduced that the expression of DNMT family might have different effects on the pathogenesis of prostate cancer. A previous report demonstrated that DNMT3a and DNMT3b had overlapping functions in global remethylation during early embryogenesis [
31]. However, the enzymes may have distinct cell- or tissue-specific functions during later embryogenesis or tumorigenesis [
32]. In addition, DNMT family expressions could potentially be controlled at the post-transcriptional level in castrated mice. Lee JY
et al. found that expression of DNMT3A and DNMT3B genes were post-transcriptionally regulated by several microRNAs, specifically
miR-1741, miR-16c, and
miR-222, and
miR-1632 via their 3’-UTR in cancerous ovaries of laying hens [
33]. In this respect, more studies are needed to prove if the observed effect on DNMTs expression is specific for prostate cancer following castration.
Moreover, we found higher methylation of global genomic DNA at the later stage after castration versus the early stage, suggesting that the expression of some important anti-oncogenes was perhaps inhibited. However, 5-Aza-CdR treatment attenuated the methylation levels in the xenograft tumors of castrated mice at both stages. Among the different groups, the methylation levels of the miR-146a promoter were positively correlated with the global methylation status and negatively associated with miR-146a expression. It was notable that the miR-146a promoter methylation levels at the CpG site of −444 bp and −433 bp, in which an NF-κB binding site located, were lower than all the other CpG sites in the combination treatment mice at the second stage. NF-κB is one of critical transcription factors for miR-146a expression, so the methylation level of NF-κB binding region in miR-146a promoter methylation is a key regular mechanism involved in miR-146a transcription. Hypermethylation of the miR-146a promoter was associated with a loss of its expression in androgen-independent cells, and altered patterns of methylation in these cancer cells might represent a form of genome instability that has been hypothesized to occur during cancer progression. The loss of mir-146a was found in high-grade AIPC tissues but not in androgen-sensitive prostate epithelium, and miR-146a was demonstrated to be a tumor-suppressor gene in modulating HA/ROCK1-mediated tumorigenicity in androgen-dependent prostate cancer [
12]. Our results suggest that the hypermethylation of the miR-146a promoter may be associated with aberrant miR-146a expression in vivo and possibly the resulting androgen insensitivity of these cells. However, to be fully understood, the precise mechanism of miR-146a action (its specific gene targets) in prostate cancer progression warrants investigation.
In conclusion, this is the first study to examine the effect of 5-Aza-CdR on the inhibition of prostate tumor growth in a castrated animal model. We also for the first time shed light on the molecular mechanisms underlying the synergistic action of 5-Aza-CdR and castration, and our findings suggest that the epigenetic regulation of miRNAs may play important roles in androgen-independent progression in patients after receiving androgen ablation therapy.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
XW performed the experiments and wrote the manuscript. HG and LR carried out animal experiments and contributed to the evaluation of treatment effects. JG and YZ participated in analyzing the data. YZ, as the corresponding author, designed the protocol and made the draft of the manuscript. All authors read and approved the final manuscript.