Cyproterone acetate enhances TRAIL-induced androgen-independent prostate cancer cell apoptosis via up-regulation of death receptor 5

DU145 (ATCC® HTB-81™), PC-3 (ATCC® CRL-1435™) and HEK293 (ATCC® CRL-1573™) cell lines were from the American Type Culture Collection (Manassas, VA). Immortalized human prostate stromal PS30 cells were kindly provided by Dr. Debra Schwinn (Duke University Medical Center) [19]. DU145 and HEK293 cells were cultured in DMEM medium with 10% fetal bovine serum (FBS); PC-3 and PS30 cells were cultured in RPMI 1640 medium with 10% FBS. Before treatment with compounds, FBS in culture medium was reduced to 2%. CPA, Bic, cycloheximide and 4-phenylbutyrate (4-PBA) were purchased from Sigma-Aldrich (St. Louis, MO). Recombinant TRAIL was a gift from Dr. Xu Luo (University of Nebraska Medical Center) [20].

The ApopNexin™ FITC Apoptosis Detection Kit (Millipore, Billerica, MA) was used for the annexin V/propidium iodide (PI) flow cytometry analysis of early apoptosis by FACSCaliber cytometry (BD Biosciences) at the Creighton University Flow Cytometry Core Facility. PARP antibody (#9532, Cell Signaling Technology, Beverly, MA) was used to detect cleavage of PARP for the western blot method of apoptosis detection. The percentage of cleaved PARP (89 kDa) in total PARP (116 kDa full length plus cleaved PARP) was used as indicator of apoptosis.

Protein extracts from cultured cells were quantified using Pierce™ Coomassie (Bradford) Protein Assay Kit (Thermo Fisher Scientific, Waltham, MA) and then subjected to western blot analysis as we reported [21]. Primary antibodies against DR5 (#AB16942, Milllipore), caspase-8 (#9746, Cell Signaling Technology), FLIP (#3210, Cell Signaling Technology), FADD (#610399, BD Transduction Laboratories™), CHOP (#2895, Cell Signaling Technology), β-actin (Santa Cruz Biotechnology) and Bid (kindly provided by Dr. Xu Luo) [20] were used to detect relevant proteins of interest.

Total RNA extraction and quantitative real-time PCR were conducted as described [22]. The DR5, DR4 and β-actin primers are listed in Additional file 1: Table S1.

All primers are shown in Additional file 1: Table S1. The DR5 promoter fragments (−711/+3) were amplified from DU145 cell genomic DNA by PCR with BglII- and HindIII-flanked primers and cloned into the pGL3-Basic luciferase reporter vector (Promega, Fitchburg, WI). Progressive deletion mutants of the DR5 promoter were created by PCR with promoter-specific primers, using the DR5 promoter luciferase plasmid (−711/+3) as a template. The DR5 luciferase plasmid DR5-P(mtCHOP) with mutations in the CHOP binding site was generated by over-lap PCR. Nucleotide changes are underlined. All constructs were validated by DNA sequencing.

DU145 cells (1.6 × 10⁵ cells) were co-transfected with 100 ng of the Renilla luciferase plasmid pRL-SV40P (Addgene) and 400 ng of firefly luciferase constructs containing the DR5 promoter region using Lipofectamine 2000 (Life Technologies). After 24 h, the cells were treated with CPA or solvent for an additional 24 h. A Dual-Glo® Luciferase Assay kit (Promega) and Sirius luciferase assay system (Berthold, Germany) were used to detect luciferase activity and the firefly luciferase activity was normalized by Renilla luciferase activity.

CHOP siRNA (#L-004819-00-0005) and control siRNA (#D-001810-01-05) were purchased from GE Dharmacon (Lafayette, CO). DU145 cells were transfected with 300 nM siRNA using Nucleofector® II Device (Lonza, Walkersville, MD) according to manufacturer’s instructions. 24 h later, cells were subjected to treatment with the indicated compounds.

Data are expressed as means ± SEM of at least three determinations. Groups were compared using Student’s t test for unpaired observations or two-way ANOVA with the Bonferroni correction for multiple comparisons. * p < 0.05; ** p < 0.01; ns, not significant.

PC-3 and DU145 are two androgen-independent prostate cancer cell lines and also often characterized as AR negative. We first examined the effect of CPA treatment on TRAIL-induced apoptosis in these two cell lines. Using the annexin V/PI assay method, we found that CPA (50 μM) alone only slightly increased apoptosis of PC-3 cells in the absence of TRAIL, while significantly enhanced TRAIL-induced cell apoptosis by 2.5-fold (Fig. 1a). Western blot analysis of PARP cleavage confirmed that CPA treatment not only increased the maximum PARP cleavage by 3-fold but also reduced the TRAIL concentration needed to induce 50% of maximal cleavage of PARP (the EC₅₀) from 50 ng/ml to 20 ng/ml (Fig. 1b). DU145 cells are highly resistant to TRAIL-induced apoptosis with only a marginal increase in PARP cleavage observed even in the presence of 100 ng/ml TRAIL. Pretreatment of CPA (50 μM) sensitized DU145 cells to TRAIL-induced apoptosis (Fig. 1c). At 100 ng/ml TRAIL, PARP cleavage was increased from 5 to 40% by CPA pretreatment.

However, another classical AR antagonist Bic (50 μM) had no effects on TRAIL induced cleavage of PARP in PC-3 and DU145 cells (Fig. 1d). As expected, immortalized normal human prostate stromal PS30 cells are resistant to TRAIL. Pretreatment with 50 μM CPA had no significant effects on TRAIL (50 ng/ml) induced cleavage of PARP in these prostate cells whereas cleavage of PARP in DU-145 cells were increased by 4-fold (Fig. 1e). Therefore, we focused our studies on DU145 cells.

The binding of TRAIL to its DR5 receptor leads to the cleavage and activation of caspase-8, a critical step in the extrinsic pathway for cell apoptosis. As shown in Fig. 2a, CPA enhanced TRAIL-induced production of the p18 fragment of caspase-8 in DU145 cells and markedly increased TRAIL-induced cleavage of the BH3-only protein Bid, a critical mediator of the mitochondrial apoptotic pathway. Pretreatment with caspase-8 inhibitor Z-IETD-FMK effectively blocked CPA/TRAIL-induced caspase-8 p18 production by over 90% (Fig. 2b). Importantly, CPA-enhanced TRAIL-induced cleavage of Bid (Fig. 2c) and PARP (Fig. 2d) was also blocked by Z-IETD-FMK pretreatment.

We next examined the components of the DISC and its negative regulator FLIP [23] in DU145 cells for illustrating the potential regulation by CPA. While FLIP and other DISC components (caspase-8 and FADD) did not show a change, DR5 protein was elevated following CPA treatment in a time- and concentration-dependent manner (Fig. 3a, b).

We further determined if the elevated DR5 protein level seen with CPA treatment was due to a prolonged DR5 protein half-life. DU145 cells without or with CPA-induced DR5 elevation were incubated with cycloheximide to block new protein biosynthesis. The disappearance of DR5 was followed over time by western blot. Figure 3c shows a representative western blot indicating that independent of CPA treatment, DR5 was gradually degraded with a half-life of at least 8 h and CPA treatment had no significant effects on DR5 degradation.

We also carried out quantitative RT-PCR analysis to investigate DR5 mRNA changes induced by CPA treatment. As shown in Fig. 3d, CPA increased DR5 mRNA expression by approximately 2-fold at 6 h and this induction was persistent over the ensuing 24 h. In addition, CPA increased DR5 mRNA levels, but not DR4 mRNA levels, in a concentration-dependent manner (Fig. 3e).

DR5 proximal promoter containing a 714-bp DNA fragment upstream of the human DR5 coding region was amplified using the genomic DNA from DU145 cells as a template, cloned into the luciferase reporter vector pGL3-Basic, and designated as DR5-P(−711/+3) (the “+3” represents the third base from the putative transcription start site). Luciferase reporter gene constructs containing the various 5’ flanking regions of the DR5 promoter were also generated (Fig. 4a, upper section). The constructs were transiently transfected into DU145 cells, and their promoter activities were determined in the presence or absence of 50 μM CPA. As shown in Fig. 4a (lower section), following deletion, the DR5-P(−552/+3) plasmid still retains the basal promoter activity, similar to that of the DR5-P(−711/+3) plasmid. CPA treatment increased the promoter activities of DR5-P(−711/+3) and DR5-P(−552/+3) by 1.9- and 2.2-fold, respectively. The DR5-P(−294/+3) plasmid still retains 62% of the DR5-P(−711/+3) basal promoter activity and displayed a similar CPA induction of DR5-P(−711/+3) (1.61- versus 1.9-fold). However, a further 96-bp deletion of DR5-P(−294/+3) resulted in a significant loss of the basal promoter activity (0.34 versus 0.62) and the complete loss of CPA induction, suggesting that this deleted 96-bp fragment (−294/−198) contains key elements of the DR5 promoter that may be responsible for the stimulatory effect of CPA.

Using TFSEARCH, a tool for locating transcription factor binding sites, we identified a consensus transcription factor CHOP binding domain at −276/-264 bp in this 96-bp fragment (Fig. 4b, upper panel). CHOP is reported to be responsible for DR5 up-regulation induced by many drugs [24]. To determine whether the CHOP-binding motif is critical for CPA-stimulated DR5 promoter activity, we mutated the consensus CHOP-binding domain in DR5-P(−552/+3) plasmid. Mutation of this consensus CHOP binding domain significantly reduced the basal promoter activity by 50% and completely abolished CPA induction (Fig. 4b, lower panel). These results demonstrate that this consensus CHOP binding domain (−276/−264) is critical for both basal and CPA-activated DR5 promoter activity.

We next examined whether CPA increased CHOP protein expression in prostate cancer cells. The basal level of CHOP protein is very low in DU145 cells, but was significantly increased following 6 h treatment with various concentrations of CPA (Fig. 5a). CPA (50 μM) treatment increased CHOP protein expression in a time-dependent manner to a maximal 6-fold induction at the time of 6 h and this induction was gradually reduced but was still 2-fold higher than the basal level at the time of 30 h (Fig. 5b). Since CHOP protein is a marker of ER stress [25], we also investigated whether CPA induced CHOP protein could be blocked by ER stress reducer 4-phenylbutyrate (4-PBA). As shown in Figs. 5c, d, after pretreatment of DU145 cells with 5 mM 4-PBA, CPA induced up-regulation of CHOP and DR5 protein were reduced by 40–50%. Furthermore, we found that neither DR5 protein expression nor TRAIL induced cleavage of PARP was affected by CPA treatment in ER stress resistant HEK293 cells (Additional file 2: Figure S1).

To further determine whether CHOP is responsible for CPA-enhanced DR5 protein expression, DU145 cells were transfected with control siRNA or CHOP-specific siRNA for 24 h, followed by CPA (50 μM) treatment for 6 h. As shown in Fig. 6a, CHOP siRNA largely abolished CPA-induced CHOP expression at the time of 6 and 30 h. Silencing CHOP expression reduced CPA-enhanced DR5 protein expression by about 50% (Fig. 6b, c). More importantly, silencing CHOP expression reduced CPA-stimulated TRAIL-induced PARP cleavage from 40 to 20% (Fig. 6b, d). An annexin V/PI flow cytometry analysis indicated that CPA treatment only increased TRAIL induced cell apoptosis by 1.3-fold in CHOP-silenced DU145 cells as compared 2-fold in control DU145 cells (Fig. 6e). Together, our results suggest that induction of CHOP plays an important role in CPA-enhanced TRAIL-induced prostate cancer cell apoptosis.