In order to reinforce DICE1 tumor suppressor function in prostate cancer, we analyzed its expression in different human androgen-dependent/-sensitive (LNCaP), androgen-independent/-insensitive (DU145, PC3, PC3-ml); and prostate cell lines isolated from prostatic primary tumor (paired normal/tumor 1532NPTX/1532CPTX cell lines) 14 . Northern blot showed that DICE-1 mRNA in all the prostate tumor cell lines tested is strongly down-regulated or undetectable compared to normal immortalized prostate 1532NPTX cells and BALB/c3T3 mouse cells transfected with the full length DICE1 cDNA (Figure 1).

DICE1 function in prostate cancer was studied by exogenously re-expressing DICE1 cDNA in the prostate tumor cell lines PC3 and DU145. This experiment was performed by transfecting the cancer cells with the pDICE1-EGFP fusion construct or pEGFP control expression plasmid (Figure 2a, left). Synthesis of a DICE1-EGFP fusion protein in PC3 (Figure 2a, right) and DU145 (data not shown) transfected cells was shown by a band around 150 kD by Western blot analysis. This size is slightly larger than the calculated size (expected ~125 kD). Similarly, a slightly larger than expected size band was previously found for human endogenous DICE1 protein functioning as integrator subunit 6 of the integrator multiprotein complex 9 .

Transfected cells were then tested for their ability to form colonies in vitro in G418 supplemented selective medium. Interestingly, we show that the re-expression of DICE1 inhibits colony formation of PC3 and DU145 cells by 50% and 90% respectively when compared to the control pEGFP-transfected cells (Figure 2b). With the androgen-dependent cell line LNCaP, previously shown to harbor missense mutation D546G in the DICE1 gene 6 , we obtained inconsistent results concerning the growth-inhibitory responses to DICE1-EGFP (data not shown).

To determine if DICE1 suppresses prostate cancer cell growth by inducing apoptosis, we analyzed its capacity to stimulate genomic DNA fragmentation in PC3 cells. PC3 cells were transiently transfected with pDICE1-EGFP or pEGFP plasmid. Subsequently, genomic DNA was extracted after 24 and 48 hours, quantified and loaded on an agarose gel. We could not detect any DNA fragmentation in PC3 transfected cells as compared to the apoptotic positive control cells (data not shown). These results were validated by loading higher amounts of genomic DNA (up to 10 μg, data not shown). In our understanding, the absence of an apoptotic signal could not be explained by a lack of sensitivity of the present assay. Indeed, we estimated approximately 30% transfection efficiency in this experiment, which is an adequate ratio to visualize DNA fragmentation. Additionally, we tested the capacity of DICE1 to induce necrosis in androgen-independent cells. As expected, this analysis performed by propidium iodide (PI) uptake assay on prostate cancer cells did not show any significant effect of DICE1 expression on the percentage of PI-positive cells (data not shown). In conclusion, our results suggest that DICE1 is inhibiting prostate cancer cell growth through an apoptosis- and necrosis-independent pathway.

To further understand the mechanisms involved in DICE1 growth inhibitory function, we examined the capacity of DICE1 to modify the cell cycle distribution of normal prostate cells RWPE-1 and prostate cancer cells LNCaP, PC3 and DU145. Similar to results obtained with normal RWPE-1 cells, re-expression of DICE1 in PC3 or DU145 cells induced a 30-50% cell reduction in the sub-G0 phase (15.1 +/- 0.52 for pEGFP-PC3 versus 10.4 +/- 0.54 for pDICE1-EGFP-PC3 and 28.4 +/- 2.21 for pEGFP-DU145 versus 14.9 -/+ 1.18 for pDICE1-EGFP-DU145), validating our hypothesis of an apoptosis-independent mechanism (Figure 3a-b). Likewise, the percentage of PC3 and DU145 cells in G0/G1 was increased about 20-30% following DICE1 re-expression suggesting a block in the G1-phase of the cell cycle (Figure 3a; 47.5 +/- 0.23 for pEGFP-PC3 versus 55.3 +/- 2.26 for pDICE1-EGFP-PC3 and 36.4 +/- 2.24 for pEGFP-DU145 versus 47.6 +/- 1.0 for pDICE1-EGFP-DU145). DU145 cells re-expressing DICE1 showed an additional 20% increase in cell number in G2/M phase emphasizing that DICE-1 could regulate the cell cycle at different levels. In agreement with the results obtained in the colony formation assay, the re-expression of DICE1 in LNCaP cell line did not present any significant effect on their cell cycle distribution.

As previously described, inhibition of the Wnt pathway in PC3 cells resulted in decreased colony formation in soft agar and in vivo tumor growth 15 . Therefore, we analyzed the expression of genes related to this pathway by PCR array. PC3 cells containing the exogenous pDICE1-EGFP fusion construct showed differences in the expression of several genes when compared with pEGFP-transfected PC3 cells indicating regulation of Wnt signaling in response to exogenous DICE1 expression (Table 1). In particular, genes like CXXC finger 4 (CXXC4), frizzled homolog 7 (FZD7) and transcription factor 7-like 1 (TCF7L1) showed an up-regulation of more than 3- to 7-fold, whereas cyclin D1 and transcription factor 7 (TCF7) were clearly down-regulated (3.8 and 3.6 fold, respectively; Table 1). To confirm the PCR array results, we repeated the expression analysis by quantitative real-time PCR for selected genes CXXC4, TCF7L1 and CCND1 in DICE1- and control-transfected PC3 and DU145 cells. The data showed a clear 3.5 fold up-regulation of TCF7L1 in PC3 and DU145 cells in response to DICE1, whereas CCND1 was down-regulated (1.7 and 5.3 fold, respectively). In the case of CXXC4, DICE1-transfected PC3 cells showed a more than 30-fold up-regulation, whereas a 7-fold down-regulation was observed in DU145 cells (Figure 4).

Human immortalized normal prostate cells RWPE-1 (American Type Culture Collection; ATCC) were grown in keratinocyte serum-free medium (Invitrogen, Carlsbad, CA) supplemented with 25 μg/ml bovine pituitary extract and 5 ng/ml epidermal growth factor. Human prostate cancer cells LNCaP, DU145, PC3 and PC3-ml were grown in RPMI 1640 medium supplemented with 10% fetal calf serum (HyClone, Perbio Science, Erembodegem-Aalst, Belgium) and 1% penicillin/streptomycin mix (Sigma, Taufkirchen, Germany). Paired cancer/normal human prostate cell lines, CPTX1532 and NPTX1532, were generated from patient undergoing radical prostatectomy and established by immortalization after micro-dissection of primary tumor cells and adjacent normal tissue respectively 14 . NPTX1532 and CPTX1532 cell lines were cultured in keratinocyte serum-free medium, 25 μg/ml bovine pituitary extract, 5 ng/ml epidermal growth factor, 2 mM L-Glutamine, 10 mM HEPES, 50 ng/ml gentamicin sulphate (all from Invitrogen, Carlsbad, CA), 5% heat inactivated fetal bovin serum (BioWhitacker, Rockville, MD), 2.5% penicillin/streptomycin mix (CellGro) and 0.5 μg/ml fungizone (CellGro). For transfection, prostate cells (1 × 10⁵) were seeded in 35 mm culture dish and incubated overnight at 37°C. The cells were then transfected for 4 hours with 4 μg pDICE1-EGFP expression plasmid 13 or pEGFP control plasmid (Clontech) by using the CLONfectin kit (Clontech) following the recommended protocol. The transfection efficiencies were evaluated by counting the EGFP-positive cells under an inverted fluorescent microscope (Axiovert 25CFL, Zeiss, Göttingen, Germany), and reporting this number to the total number of cells (regular light).

BALB/c3T3 mouse cells were infected with retroviral particles containing either the full length DICE1 cDNA or no insert. The retroviral packaging line Phoenix (provided by Dr. Gary Nolan at Stanford University) was used for viral transduction studies 23 . Cells were selected in 1 μg/ml puromycin for 3 days. Total RNA was isolated following the guanidinium isothiocyanate method of Chomczynski and Sacchi 24 . Northern blot analysis was carried out using 10 mg of total RNA in glyoxal agarose gels. Size fractionated RNA was transferred to nylon filters electrophoretically in 1× Tris Acetate EDTA buffer (TAE). Probes were labeled by random priming and hybridization was carried out using a solution of 7% SDS, 0.25 M Na₃PO₄, 5.6 mM Na₄P₂PO₇, 2 mM EDTA. Washing was carried out under standard conditions.

After transfection, prostate cells were incubated over 2-3 weeks in the presence of G418 antibiotic (Sigma) to allow colonies to develop. At the endpoint of the experiment, the medium was removed, the colonies washed in PBS and stained with 0.5% crystal violet (Sigma) in methanol. Each colony formation assay was carried out in triplicate and repeated at least three times. The working concentration of G418 (380 μg/ml for PC3 and 240 μg/ml for DU145 cells) was defined as the lowest dose of antibiotic that kills 100% of non-transfected cells in 5-7 days from the start of G418 selection.

Cell extracts of EGFP and DICE1-EGFP transfected cells were produced by RIPA lysis and separated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) according to Harlow and Lane 25 . The gel was blotted on to a PVDF transfer membrane (PVDF Polyscreen, NEN Life Science Products, Boston, MA, USA) using a semi-dry blotting device (Trans-blot SD, BioRad, München, Germany). Immunodetection was carried out by using affinity-purified monoclonal anti-GFP antibody JL-8 (Living Colors, Clontech, Palo Alto, U.S.A) and chemiluminescence as described in 26 .

For cell cycle analysis, prostate cells were trypsinized and washed with ice-cold PBS. Then cells were fixed with 70% ice-cold ethanol for 1 hour, followed by incubation in freshly prepared nuclei staining buffer (200 μg/ml RNase plus 20 μg/ml Propidium Iodide-PI in PBS) for 1 hour at 37°C. Cell cycle histograms were generated after analysis of PI-stained cells by fluorescence-activated cell sorting (FACS) with a Becton Dickinson FACVantage SE Cell Sorter. For each sample, triplicates were performed and >1 × 10⁴ events were recorded. Histograms generated by FACS were analyzed by Cell Quest Software to determine the percentage of cells in each phase (G1-S-G2/M). Statistical evaluation of the data was done by paired Student's t test using the SPSS 11.5 software for Windows. SE and P values are shown where appropriate.

Total RNA was isolated by using the RNeasy Mini kit (Qiagen GmBH, Hilden, Germany), following the manufacturer's recommendations. RNA (2.5 μg) of PC3 and DU145 cells were converted to cDNA by SuperScript III Reverse Transcriptase, using a mixture of Oligo dT (0.25 μg) and Random Primer (0.5 μg; all from Invitrogen, Carlsbad, CA, USA) in a volume of 11 μl RNase-free water. Gene expression (Table 1) was quantified in triplicate with the iCycler and MyIQ amplifiers (Bio-Rad) using a SYBR-Green Supermix kit as recommended by the supplier (Bio-Rad). The real-time PCR was performed with primers and annealing temperatures described in table 2; expression of each of the INTS6/DICE1 target genes was quantified in three distinct batches of transfection.

The expression of 84 genes related to human Wnt signaling pathway was analyzed in PC3 cells by quantitative real-time PCR using RT² Profiler PCR Array technology (SABiosciences, Frederick, MD, USA). The provided Master Mix, containing dNTP's, polymerase, MgCl₂ and SYBR-Green, and the cDNA were used as described by the manufacturer. The RT-PCR reaction was performed in the iCycler (Bio-Rad). For data analysis Ct values were used for the gene of interest and the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (G3PDH). The relative quantification of gene expression was determined using the 2^-ΔΔCT method as described by Livak and Schmittgen 27 .