Expression of the fat-1 gene diminishes prostate cancer growth in vivo through enhancing apoptosis and inhibiting GSK-3β phosphorylation

Prostate cancer is the most common cancer and the second leading cause of cancer-related death in the United States (1). Dietary fat as a potential risk factor for cancer has been the focus of many epidemiologic clinicians and basic researchers, but the findings have been inconclusive (2–7). There are two types of polyunsaturated fatty acids (PUFA), ω-3 fatty acids (ω-3 or n-3 PUFA) and ω-6 fatty acids (ω-6 or n-6 PUFA), which are the essential fatty acids that cannot be synthesized by mammals. They are derived entirely from the diet (8, 9). As the constituents of the cell membrane, n-6 and n-3 PUFAs serve as precursors of eicosanoids that mediate both preventive and beneficial effects. It has been reported that the ratio of n-6 to n-3 fatty acids in today's western diet is 10-30:1 instead of 1:1, the characteristics of the traditional diet of which human beings have consumed for many millennia (10). This unbalanced n-6/n-3 ratio may contribute to the increased incidence of “modern” diseases, which include various types of cancer and diabetes (10).

Numerous studies support that the ratio of n-6/n-3 PUFAs affects tumorigenesis (11, 12). The inhibitory effects of n-3 PUFA-rich diets have been reported in mouse models of various tumor types, including prostate cancer (13–17). In a chemical-induced colon cancer animal model, diets with high levels of n-3 PUFA reduced the incidence, multiplicity, and size of colon tumors compared with diets rich in n-6 PUFAs (18). In a prostate-specific PTEN knockout mouse model, diets with high levels of n-3 PUFAs reduced prostate cancer growth, slowed histopathologic progression, and increased animal survival, whereas n-6 PUFA-rich diet had opposite effects (19). In that study, the fat-1 gene, cloned from Caenorhabditis elegans that encodes a n-3 fatty acid desaturase, which converts n-6 PUFA to n-3 PUFA, was introduced into the PTEN knockout mice. The fat-1 gene expression in the host reduced the spontaneous prostate cancer growth similar to the effects of a n-3 PUFA-rich diet (19). In another study, mouse melanoma B16 cells were implanted into fat-1 transgenic mice (15). A dramatic reduction of both melanoma formation and growth in fat-1 transgenic mice was reported. Taken together, these findings suggest that introduction of the fat-1 gene into the host, or possibly into tumor cells, could be used for cancer treatment. Indeed, it was reported that adenoviral transfer of the fat-1 gene into breast cancer cells and lung cancer cells had an antitumor effect in vitro (12, 20). However, the effects of the fat-1 gene expression on tumor growth in vivo have not been studied and the mechanisms remain unclear. Accordingly, we first investigated the inhibitory effects of the fat-1 gene expression on prostate cancer DU145 and PC3 cell proliferation in vitro. Then, we examined the effects of the fat-1 gene expression on prostate cancer growth in vivo. Finally, we evaluated the signaling pathways that may be potential mechanisms for the effects of the fat-1 gene transfer into the tumor cells.

The Ki-67 monoclonal antibody was purchased from Invitrogen. The ApopTag Peroxidase In situ Apoptosis Detection kit was purchased from Millipore. Antibodies against caspase-3, caspase-9, GSK-3β, phosphorylated GSK-3β, β-catenin, Akt, and cyclin D1 were purchased from Cell Signaling Technology. The Bcl-2 antibody was purchased from R&D Systems. The phosphorylated Akt antibody was purchased from Santa Cruz Biotechnology. The caspase-3 activity assay kit was purchased from EMD Biosciences. Arachidonic acid (AA), docosahexaenoic acid (DHA), and eicosapentaenoic acid (EPA) were purchased from Cayman Chemical. All of the other chemical reagents were purchased from Sigma.

The coding region of the fat-1 gene from C. elegans (GenBank: NM_001028389) was optimized for mammalian cell expression and named as mfat-1. The mfat-1 cDNA was synthesized and constructed into the mfat-1 expression vector, pST180, consisting a cytomegalovirus enhancer with a chicken β-actin promoter and a pgk-neo expression cassette as a selection marker. The control construct, pST128, contained a SV40-neo expression cassette.

Human prostate cancer cell lines DU145 and PC3 and nontumorigenic human prostate epithelial cells RWPE-1 were obtained from the American Type Tissue Collection. The DU145 and PC3 cells were cultured in RPMI 1640 (Invitrogen) supplemented with 10% fetal bovine serum, 100 units/mL penicillin, and 100 μg/mL streptomycin (Invitrogen). The RWPE-1 cells were grown in keratinocyte/serum-free medium supplemented with 5 ng/mL human recombinant epidermal growth factor and 0.05 mg/mL bovine pituitary extract (Invitrogen). All cells were maintained in 10 cm tissue culture dishes in a 37°C incubator equilibrated with 5% CO₂ in humidified air.

DU145 and PC3 cells were transfected with linearized pST180 or pST128 by electroporation and selected with G418 at 200 to 400 μg/mL. The G418-resistant colonies were pooled after 2 weeks of cell clone selection to limit possible variations among different clones. The mfat-1 transgene expression and its function were confirmed by reverse transcription-PCR (RT-PCR) and fatty acid profile analysis.

Total RNA was extracted from parental prostate cancer DU145 and PC3 cells, and DU145 and PC3 cells were transfected with either pST180 or pST128, using TRIzol reagent (Invitrogen), then subjected to RT-PCR for detection of mfat-1 mRNA expression. The PCR primers used for mfat-1 consisted of sense 5′-GGACCTTGGTGAAGAGCATCCG-3′ and antisense 5′-GCGTCGCAGAAGCCAAAC-3′ resulting in a PCR product of 438 bp. The PCR primers for glyceraldehyde 3-phosphate dehydrogenase (as control) consisted of sense 5′-CCATGGAGAAGGCTGGGG-3′ and antisense 5′-CAAAGTTGTCATGGATGACC-3′ resulting in a PCR product of 194 bp. RT-PCR was done with 1 μg total RNA using the Access RT-PCR system (Promega) in a thermal cycler (GeneAmp PCR System 2700; Applied Biosystems) under the following conditions: first-strand cDNA was synthesized at 48°C for 45 min and then denatured at 94°C for 2 min for the first cycle and at 45 s for additional 30 cycles; annealing was done at 60°C for 45 s and extension at 68°C for 2 min. Final extension was at 68°C for 5 min. The PCR products were subjected to electrophoresis on a 1.5% agarose gel and stained with ethidium bromide.

Lipids were extracted as in a previous report (8). Briefly, the samples were homogenized in a mixture of methanol, chloroform, and water. After 15 min, chloroform was added and the samples were vortexed and centrifuged. The lower phase was dried under nitrogen and resuspended in boron trifluoride methanol. The samples were heated at 90°C for 30 min and extracted with 4.0 mL pentane and 1.5 mL water. The mixtures were vortexed and the upper phase was recovered. The extracts were dried, resuspended, and injected into a capillary column (SP-2380, 105 m × 53 mm ID, 0.20 μm film thickness; Sigma). The gas chromatography was done on a Perkin-Elmer Clarus 500. Identification of components was done by comparison of retention times with those of authentic standards (Sigma).

Cell proliferation was measured with the use of a CellTiter 96 AQueous Non-Radioactive Cell Proliferation Assay (Promega). Briefly, prostate cancer DU145 and PC3 cells, cells transfected with either pST180 or pST128 vector, and normal prostate epithelial cells (RWPE-1) were plated in 96-well plates at a density of 3,000 per well in 200 μL appropriate growth medium overnight. The medium was changed to RPMI 1640 plus 0.5% fetal bovine serum with AA (10-100 μmol/L), DHA (10-100 μmol/L), or EPA (10-100 μmol/L). The cells were incubated at 37°C in a humidified 5% CO₂ atmosphere for 24, 48, or 72 h, and 20 μL combined MTS/PMS solution was added. After incubation for 2 h at 37°C, the absorbance of each well was recorded at 490 nm using an ELISA plate reader.

Six-week-old male severe combined immunodeficient (SCID) mice (Charles River Laboratories) were housed under pathogen-free conditions in accordance with the NIH guidelines using an animal protocol approved by the University of Pittsburgh Animal Care and Use Committee. Xenograft tumors were established as described previously (21). Briefly, single-cell suspensions (1 × 10⁶ cells) of the prostate cancer DU145 or PC3 cells that were transfected with either the fat-1 gene or the control vector in RPMI 1640 were prepared and mixed 1:1 with Matrigel (Collaborative Biomedical Products). The SCID mice were injected s.c. into the right flank at 100 μL/site using a 23-gauge needle. S.c. tumor growth was monitored by palpation, and two perpendicular axes were measured using an electronic caliper; the tumor volume was calculated using the formula as described previously (22): volume = length × width² / 2. Mice were assigned to one of the four groups (n = 10 per group) based on the tumor cells injected: (a) DU145 pST128, (b) DU145 pST180, (c) PC3 pST128, and (d) PC3 pST180 cells. The tumors were measured twice weekly and allowed to grow for 6 weeks. On the completion of this study, the mice were sacrificed. Tumor tissues, tibia, and vertebra as well as blood serum were collected for further analysis.

Xenograft tumors were harvested and placed in 10% formalin, embedded in paraffin, and sectioned at 5 μm thickness. Sections were examined using standard H&E staining for routine histology. To evaluate the proliferation of xenografts, sections were deparaffinized, rehydrated, and stained with Ki-67 monoclonal antibody following a modified protocol (23). To evaluate apoptosis, the similar sections with proliferation sections were subjected to terminal deoxynucleotidyl transferase-mediated nick end labeling analysis using the ApopTag Peroxidase In situ Apoptosis Detection kit (Millipore) according to the manufacturer's directions. Both the Ki-67 labeling index and the apoptotic index were calculated as the percentage of positive tumor nuclei divided by the total number of tumor cells examined. At least 1,000 tumor cells per specimen were examined in five randomly selected fields by light microscopy (×400) by an investigator who was blinded to the animal groups as we have described previously (24).

Cell lysates from parental prostate cancer DU145 and PC3 cells and the cells that were transfected with the fat-1 gene or the control vector were prepared using standard procedures. All the samples were measured for total protein concentration using a BCA assay (Pierce) to ensure equal loading. Loading buffer was added to 40 μg protein, and samples were boiled before being resolved on 12% SDS-PAGE gels. Then, the samples were transferred onto polyvinylidene difluoride membranes (Bio-Rad). The blots were blocked using blocking reagents at room temperature for 1 h with shaking followed by incubation overnight with primary antibody for GSK-3β, phosphorylated GSK-3β, β-catenin, caspase-3, caspase-9, Akt, phosphorylated Akt, Bcl-2, or cyclin D1 (at appropriate antibody dilution in primary antibody dilution buffer). The blots were washed and then incubated for 1 h with either anti-rabbit or anti-mouse IgG-horseradish peroxidase (1:5,000). After washing, the bands were detected using enhanced chemiluminescence (Amersham Biosciences) and exposed to light-sensitive film. As a control for equal loading of the proteins, immunoblots for glyceraldehyde 3-phosphate dehydrogenase (Santa Cruz Biotechnology) were done on the stripped membranes.

Caspase-3 activity was determined with a caspase-3 activity assay kit (EMD Biosciences) according to the manufacturer's protocol. Briefly, DU145 and PC3 cells that were transfected with either the fat-1 gene expression plasmid or the control vector were lysed with lysis buffer. RWPE-1, DU145, and PC3 cells were lysed after treatment with AA (30 μmol/L), DHA (30 μmol/L), camptothecin (15 μmol/L; a known apoptosis inducer), or vehicle control. Caspase-3 substrate I (Ac-DEVD-ρNA) was added to cell lysates and the mixtures were incubated for 2 h in a humidified environment at 37°C. Before the DEVD-ρNA cleavage, cell lysates were preincubated for 10 min at 37°C either with or without 0.1 μmol/L caspase-3 inhibitor I (Ac-DEVD-CHO). The concentration of the ρNA released from the caspase-3 substrate was measured from the absorbance values at 405 nm, and the quantity of ρNA was calculated from a calibration curve of ρNA standards.

The in vitro invasion assay was done using prostate cancer DU145 or PC3 cells that were transfected with either the fat-1 gene or the control vector. The invasiveness of cells was evaluated in 24-well cell invasion kit from Millipore as directed by the manufacturer. Briefly, the upper and lower culture compartments of each well were separated by polycarbonate membranes (8 μm pore size). The membranes in some wells were precoated with 100 μg/cm² collagen matrix (Matrigel). To assess the ability of the cells to penetrate the Matrigel, the cells (2 × 10⁵) in 0.3 mL serum-free RPMI 1640 were placed in the upper compartment of wells. In the lower chamber, point 5 mL RPMI 1640 containing 10% fetal bovine serum was added. The Transwell chambers were incubated for 24 h at 37°C in 95% air and 5% CO₂. Cell penetration through the membrane was detected by staining the cells on the porous membrane and quantified by dissolving stained cells in 10% acetic acid and transfer a consistent amount of the dye/solute mixture into a 96-well plate for colorimetric reading. The absorbance was recorded at 560 nm.

Statistical analysis was done using Statview software (Abacus Concepts). ANOVA was used for initial analyses followed by Fisher's protected least significant difference for post hoc analyses. Differences with a P < 0.05 were determined as statistically significant.

To first determine the effects of the fat-1 gene expression on prostate cancer cell proliferation in vitro, prostate cancer DU145 and PC3 cells were transfected with the fat-1 gene (pST180) or the control vector (pST128). RT-PCR with fat-1-specific primers revealed that the prostate cancer DU145 and PC3 cells that were transfected with the fat-1 gene successfully expressed the fat-1 gene (Fig. 1A). As expected, the parental cells and the cells that were transfected with the control vector did not express fat-1 mRNA. To verify if the transfected fat-1 gene was functional, gas chromatography was done to test the composition of PUFAs in these cells. The production of n-3 PUFAs was significantly increased in the cells that were transfected with the fat-1 gene compared with the cells that were transfected with the control vector (Fig. 1B). The beneficial alteration of the PUFA composition of total cellular lipids in the cells transfected with either the fat-1 gene (pST180) or the control vector (pST128) is listed in Table 1. The ratios of n-6/n-3 PUFAs were 5.36 and 7.36 in DU145 and PC3 cells that were transfected with the control vector compared with 1.55 and 0.44 in DU145 and PC3 cells that were transfected with the fat-1 gene, respectively (P < 0.001). Transfection of the control vector did not change the fatty acid profile in DU145 and PC3 cells compared with the parental controls (data not shown). To further determine whether the beneficial alteration of PUFAs in prostate cancer cells inhibits tumor cell proliferation in vitro, the cells were cultured for different time points, and a cell proliferation assay was done (Fig. 1C). Expression of the fat-1 gene significantly inhibited prostate cancer cell proliferation in the cells that were transfected with the fat-1 gene compared with the cells that were transfected with the control vector.

To evaluate whether fat-1 gene expression in prostate cancer cells alter prostate cancer cell growth in vivo, the pST180- or pST128-transfected prostate cancer DU145 and PC3 cells were s.c. implanted into SCID mice for 6 weeks. The tumor volume was monitored twice weekly by caliper measurement. Results in Fig. 2A show that fat-1 gene expression significantly diminished tumor volume compared with the cells that were transfected with the control vector. Interestingly, we failed to observe tumor growth from the pST180-transfected PC3 cells in 3 of 10 animals. To further determine the inhibitory effects of fat-1 gene expression, we did immunohistochemical staining of Ki-67, a proliferation marker, in the tumor tissues. We found that Ki-67-positive nuclei were significantly decreased in the tumor cells that were transfected with the fat-1 gene compared with the cells that were transfected with the control vector (Fig. 2B). To determine the effects of fat-1 gene expression on apoptosis, ApopTag staining was done. Immunohistochemical staining revealed that apoptotic nuclei were significantly elevated in the tumor cells that were transfected with the fat-1 gene compared with the cells that were transfected with the control vector (Fig. 2C).

To elucidate the potential mechanisms of the inhibitory effect of the fat-1 gene expression on the tumor cell growth, the parental DU145 and PC3 cells were treated in vitro with n-3 PUFA DHA and EPA, n-6 PUFA AA, or vehicle control for 48 h. Both DHA and EPA, but not AA, decreased prostate cancer cell viability measured by a cell proliferation assay (Fig. 3A). In the nontumorigenic human prostate epithelial RWPE-1 cells, neither DHA nor EPA affected cell viability, which suggests that the apoptotic effect of DHA and EPA is neoplastic cell specific. Using DU145 cells as an example, DHA dose-dependently inhibited DU145 cell proliferation in vitro (Fig. 3B). We further examined the effect of DHA on caspase-3, caspase-9, and Bcl-2 expression. DHA, but not AA, induced caspase-3 and caspase-9 cleavage, reduced Bcl-2 expression (Fig. 3C), and induced caspase-3 activity in DU145 cells (Fig. 3D) as well as in PC3 cells (data not shown). As expected, expression of the fat-1 gene also induced caspase-3 activity in DU145 cells (Fig. 3D). A specific caspase-3 inhibitor significantly decreased DHA-induced caspase-3 activity in these cells.

To further investigate the potential mechanisms of the fat-1 gene expression on growth inhibition of tumor cells, we examined the protein expression of GSK-3β, phosphorylated GSK-3β, β-catenin, cyclin D1, Akt, and phosphorylated Akt (Fig. 4). We found that the fat-1 gene expression reduced GSK-3β phosphorylation and downstream protein expressions of β-catenin and cyclin D1. As a parallel study, the parental DU145 cells were treated with DHA. We observed the same pattern of inhibition of GSK-3β phosphorylation and β-catenin and cyclin D1 expressions. We did not observe alterations on Akt expression and Akt phosphorylation.

To explore other possible mechanisms of the fat-1 gene expression on prostate cancer development, an in vitro cell invasion assay was done. Prostate cancer DU145 and PC3 cells that were transfected with the fat-1 gene significantly decreased cell invasive ability compared with the cells that were transfected with the control vector (Fig. 5). As expected, we did not observe significant differences on cell invasion between the parental tumor cells and the cells transfected with the control vector (data not shown). These results suggest that, in addition to the inhibitory effects on the tumor cell proliferation, the fat-1 gene transfer also reduces prostate cancer tumor cell invasion.

In the current study, we first transfected the fat-1 gene directly into prostate cancer DU145 and PC3 cells. We confirmed the inhibitory effects of fat-1 gene expression on prostate cancer cell proliferation in vitro. This was achieved by transfection of the C. elegans fat-1 gene, encoding a n-3 fatty acid desaturase, which converts n-6 to n-3 PUFA, into tumor cells. To test the feasibility of using the fat-1 transgene into prostate cancer cells as a potential gene therapeutic approach, we implanted the DU145 and PC3 cells that were transfected with the fat-1 gene or the control vector into SCID mice. The tumor volume was monitored for 6 weeks. We observed that, in vivo, the fat-1 gene expression in both DU145 and PC3 cells diminished tumor growth via induction of apoptosis and inhibition of cell proliferation. These results provide evidence, for the first time, that fat-1 gene transfer directly into the tumor cells could be used as a novel therapeutic approach for prostate cancer treatment.

To explore the potential mechanisms of antitumor effects of fat-1 gene expression in prostate cancer cells, DU145 and PC3 cells were treated with DHA, EPA, or AA. We found that the DHA and EPA induced prostate cancer cell apoptosis, mediated by cleavage of caspase-3 and caspase-9 and decrease of Bcl-2 expression. Importantly, we observed that neither DHA nor EPA affected the cell viability in a nontumorigenic human prostate epithelial cell line, indicating that not only the tumor cells were targeted but also the apoptotic effect of DHA and EPA might have no harmful effects on normal cells. We further found that fat-1 expression in the tumor cells reduced GSK-3β phosphorylation and downstream protein expression of β-catenin and cyclin D1. Because GSK-3β is an important negative regulator for Wnt signaling pathways (25), we postulate that GSK-3β may play a key role in this action of fat-1. In support of this finding, it was recently reported that n-3 PUFAs block tumor cell growth at least in part through inhibition of Wnt/β-catenin and cyclo-oxygenase-2 signaling pathways (26). We are aware that phosphatidylinositol 3-kinase/Akt pathway has been reported to modulate β-catenin activity through GSK-3β (27, 28). However, in this study, we did not observe alterations on Akt expression and Akt phosphorylation by the fat-1 gene expression in either the DHA-treated tumor cells or the DHA-treated parental cells. This finding suggests the existence of Akt-independent regulation of GSK-3β in prostate cancer cells.

To further determine whether other mechanisms may exist for the inhibitory function of fat-1 gene on prostate cancer development, we measured alterations in the tumor cell invasive ability after the fat-1 gene transfer using an in vitro cell invasion assay. Both DU145 and PC3 cells that were transfected with the fat-1 gene significantly decreased cell invasion compared with the cells that were transfected with the control vector. In support of our findings, it was reported that decreased n-6/n-3 ratio reduced the invasive potential of human lung cancer cells by down-regulation of cell invasion-related genes (12). Therefore, our findings from this study may extend the beneficial effects of the fat-1 gene transfer.

We are also aware that enzymatic conversion of fatty acids also occurs extracellularly in the microenvironment. It was recently reported that melanoma growth was reduced in fat-1 transgenic mice, suggesting that the tumor growth microenvironment plays a critical role in tumor development (15). It was also reported that endogenously increased tissue levels of n-3 PUFA in fat-1 transgenic mice had both lower incidence and growth rate of colon tumors induced by inflammation. This was accompanied by lower activity of nuclear factor-κB, higher expression of transforming growth factor-β in the colons, and lower expression of inducible nitric oxide synthase in the tumors of fat-1 animals (18). These results suggest a mechanism by which n-3 PUFA suppresses tumorigenesis through dampening of inflammation and nuclear factor-κB activity.

Overall, this is the first exploration of the function of the fat-1 gene in tumor cells in vivo. We have shown that the fat-1 gene expression inhibits tumor cell proliferation and induces tumor cell apoptosis. This action is partially through Wnt/GSK-3β/β-catenin signaling pathway. The conversion of n-6 to n-3 PUFA through transfer of the fat-1 gene into tumor cells was shown as an effective therapeutic approach for prostate cancer treatment.

No potential conflicts of interest were disclosed.

We thank Drs. G David Roodman and Evan T. Keller for helpful discussions, Dr. Angela VanDyke, for technical help, and Sarabeth Borowiec for editing.