PLZF Mediates the PTEN/AKT/FOXO3a Signaling in Suppression of Prostate Tumorigenesis

Promyelocytic leukemia zinc finger (PLZF) protein expression is closely related to the progression of human cancers, including prostate cancer (PCa). However, the according context of a signaling pathway for PLZF to suppress prostate tumorigenesis remains greatly unknown. Here we report that PLZF is a downstream mediator of the PTEN signaling pathway in PCa. We found that PLZF expression is closely correlated with PTEN expression in a cohort of prostate cancer specimens. Interestingly, both PTEN rescue and phosphoinositide 3-kinase (PI3K) inhibitor LY294002 treatment increase the PLZF expression in prostate cancer cell lines. Further, luciferase reporter assay and chromatin immunoprecipitation assay demonstrate that FOXO3a, a transcriptional factor phosphorylated by PI3K/AKT, could directly bind to the promoter of PLZF gene. These results indicate that PTEN regulates PLZF expression by AKT/FOXO3a. Moreover, our animal experiments also demonstrate that PLZF is capable of inhibiting prostate tumorigenesis in vivo. Taken together, our study defines a PTEN/PLZF pathway and would shed new lights for developing therapeutic strategy of prostate cancer.


Introduction
Prostate cancer (PCa) is the most frequently occurred cancer and the second most common cause of cancer mortality of male [1], whose progression and pathogenesis have not been well studied. Phosphatase and tensin homolog deleted on chromosome 10 (PTEN), a classic tumor suppressor, is frequently mutated or deleted in human prostate cancer and is associated with advanced prostate cancer [2,3,4]. Animal models also demonstrated that the loss of PTEN is strongly correlated to prostate cancer initiation and malignancy [5,6]. PTEN is best known as an antagonist of phosphoinositide 3-kinase (PI3K)/AKT signaling pathway, a major component of the pathological avenues for a wide range of cancers development, including prostate cancer [7]. The PI3K/ AKT signaling is involved in the regulation of cellular proliferation and survival [8]. Uncontrolled activation of this pathway, mainly caused by the loss-of-function of PTEN, accounts for cellular transformation and cancer development in prostate [7]. However, the detailed mechanisms at transcriptional levels remain largely unknown. We recently knocked out PTEN in spermatogonial stem cells (SSCs) and observed tumorigenesis of SSCs (see Figure S1) associated with loss of SSC marker promyelocytic leukemia zinc finger protein (PLZF) (data not shown). Besides this, recent studies also suggested that PLZF expression is regulated by PTEN activation [9,10]. Based on these findings, we thought that PLZF might be involved in the PTEN signaling in the regulation of prostate tumorigenesis. Interestingly, we found that indeed PLZF expression is positively correlated with PTEN in human PCa samples and can be influenced by the treatment of PI3K inhibitor and AKT activation in PCa cell lines.
To understand the signaling mechanism how PLZF is regulated by the PTEN/PI3K/AKT pathway, we analyzed PLZF promoter to find its potential transcriptional regulators coupled to the PI3K/ AKT pathway. We identified seven putative binding sites for forkhead transcription factor FOXO3a, which is one of the downstream targets of AKT [11]. Phosphorylation of FOXO3a by activated AKT promotes its nucleic export, thus blocking the binding of FOXO3a to the promoter of its downstream genes, which ultimately impeded its regulation of cell proliferation and survival [12]. It has been demonstrated that FOXO3a, the most highly expressed FOXO family members in prostate cancer cells, is largely deregulated with the progression of the disease [13]. Our further experiments confirmed that PLZF promoter can be regulated by FOXO3a. Thus, PLZF is a new target of the PTEN/PI3K/AKT/FOXO3a signaling. Furthermore, our animal experiments also demonstrated that PLZF is capable of inhibiting tumorigenesis of a PTEN-null PCa cell line. Taken together, these findings indicate that PLZF works at the downstream of the PTEN signaling in suppression of PCa tumorigenesis.

PLZF expression is positively correlated with PTEN in prostate cancer
Since our previous study has indicated that PLZF expression may be correlated with PTEN in spermatogonial stem cells, we attempted to confirm the correlation in prostate cancer. For this purpose, we first performed immunohistochemical staining to detect the expression pattern of PTEN and PLZF of 50 prostate cancer specimens, which were divided into low, moderate, and high grade of subgroups according to Gleason score (GS) ranging from 5 to 9 (Table S1), as described in Materials and methods. The representative photomicrographs from 6 cases of prostate cancer for PTEN and PLZF staining are shown in Figure 1A. In each subgroup, PLZF staining was relatively weaker when PTEN staining was low (PTEN low ), whereas PLZF staining was more intense when PTEN staining was high (PTEN high ) ( Figure 1A). Moreover, the staining for PTEN and PLZF were gradually reduced with increased Gleason score of prostate cancer specimens ( Figure 1A). We next analyzed the relationship between PTEN/PLZF expression and the development of prostate cancer. Overall, the low-grade group had significantly higher PTEN and PLZF scores than the moderategrade and the high-grade subgroups (Figure 1B/C), suggesting that PLZF may also play a role in the clinical progress of prostate cancer like PTEN. Of great interests, Pearson's chisquare test showed that there was a remarkable positive correlation between PTEN and PLZF expressions ( Figure 1D/ E). A further analysis of the correlation of PTEN and PLZF was summarized in Table 1, in which most cases exhibited higher PTEN and PLZF expressions in low-grade cancer specimens, whereas most cases exhibited lower PTEN and PLZF expressions in moderate-and high-grade cancer specimens (Table 1). All these data indicate a cooperativity of PTEN and PLZF in prostate tumorigenesis.

PTEN/PI3K signaling regulates PLZF expression in prostate cancer cells
The positive correlation between PTEN and PLZF expressions in prostate cancer tissues promoted us to speculate that PTEN regulates PLZF expression. We firstly transfected PTEN into PC3 cells and LNCaP cells, which are absent of endogenous PTEN protein expression. As expected, PTEN overexpression inhibited AKT activation and it increased PLZF expression accordingly ( Figure 2A). Since PTEN is known as a lipid phosphatase that antagonizes the PI3K/AKT signaling pathway [7], we next treated prostate cancer cell lines PC3 and/or LNCaP cells with PI3K inhibitor LY294002 for hours as indicated. The results showed that LY294002 treatment significantly inhibited phosphorylation of AKT ( Figure 2B), suggesting its effectiveness to inhibit PI3K/AKT activity. More interestingly, quantitative real-time PCR analysis showed that LY294002 treatment increased PLZF expression in a timedependent manner in PC3 cells ( Figure 2C). In line with this, LY294002 also increased PLZF protein in PC3 and LNCaP cells ( Figure 2D). All these data propose that PTEN upregulates PLZF expression through inhibiting PI3K/AKT signaling.

FOXO3a regulates PLZF expression in prostate cancer cells
Next we addressed how PTEN/PI3K/AKT regulates PLZF expression. It has been exhibited that FOXO3a, which is the most highly expressed FOXO family member in prostate cancer cells, is closely related with the progression of prostate cancer [13]. AKT was also reported to phosphorylate FOXO3a at three conserved sites (Thr32/S253/S315), thus sequestering FOXO3a in the cytoplasm from its transcriptional involvement in the nucleus [14]. Indeed, here we showed that ectopic expression of constitutivelyactivated AKT (AKT CA ) increased, while dominant-negative form of AKT (AKT DN ) decreased the phosphorylations of FOXO3a ( Figure 3A). Also, AKT CA expressions sequestered FOXO3a in the cytoplasm, while AKT DN expression kept it in the nuclei of PC3 cells ( Figure 3B). Accordingly, PTEN overexpression inhibited the phosphorylations of FOXO3a (Figure 2A). Thus, we transfected wild-type FOXO3a or its mutant (FOXO3a TM ) at three conserved AKT-phosphorylated sites into PC3 cells for 24 hours. Immunofluroscent staining exhibited that wild-type FOXO3a protein was scattered in the cytoplasm and FOXO3a TM was accumulated in the nucleus ( Figure 3C). More intriguingly, FOXO3a TM expression significantly elevated PLZF expression ( Figure 3C). Further, quantitative real-time PCR and western blot analysis demonstrated that FOXO3a TM sufficiently induced the expression of PLZF ( Figure 3D/E). On the other hand, the knockdown of FOXO3a by its specific shRNA (shFOXO3a-1 and shFOXO3a-2) significantly reduced the expression of PLZF in LNCaP cells ( Figure 3F). All these results strongly suggest that FOXO3a may be a mediator of the PTEN/PI3K/AKT signaling to regulate the expression of PLZF.

FOXO3a directly binds to the promoter of PLZF gene
To further determine if FOXO3a regulates PLZF expression by its transcriptional function, we conducted luciferase reporter assay. Bioinformatic analysis found seven potential forkhead response elements (FHRE) within a 2.3 kb region containing the 59 upstream sequences and the first exon of PLZF gene ( Figure 4A). We subcloned the DNA fragment into the luciferase reporter vector pGL3-basic, which was co-transfected with empty vector or pcDNA3.1(+)-FOXO3a TM into PC3 cells. The results showed that FOXO3a TM dramatically increased the luciferase activity ( Figure 4A). Furthermore, we constructed luciferase reporters driven by different FHREs ( Figure 4B), which were co-transfected together with FOXO3a TM into PC3 cells. As shown in Figure 4B, FHRE1 and FHRE2 were sufficient for FOXO3a TM -induced luciferase expression, which could be abrogated by mutations on FHRE1, indicating that FHRE1 is indispensable for the transcription of PLZF driven by FOXO3a. The chromatin immunoprecipitiation (ChIP) results also supported that FOXO3a directly bound to the promoter of PLZF in the region of FHRE1 and FHRE2 ( Figure 4C). Given all these data, we conclude that FOXO3a upregulates PLZF expression by its transcriptional activity.

FOXO3a expression is positively correlated with PLZF in prostate cancer
Given that we have demonstrated that PTEN upregulated PLZF through FOXO3a by its transcriptional activity, it needs to be further confirmed whether this correlation also exists in the clinical specimens. As above, we detect the expression pattern of FOXO3a in the prostate cancer specimens by immunohistochemistry staining (Table S1). The representative photomicrographs from 3 cases of prostate cancer for FOXO3a staining are shown in Figure 5A. The staining for FOXO3a was reduced with increased Gleason score of prostate cancer specimens ( Figure 5A). Next we analyzed the correlation between FOXO3a expression and the development of prostate cancer. Overall, the low-grade group had significantly higher FOXO3a scores than the moderate-grade and the high-grade subgroups ( Figure 5B), suggesting that FOXO3a may gradually lose its function along with the progression of prostate cancer. As expected, we presented a significant positive correlation between FOXO3a and PLZF expressions by Spearman's Correlation analysis ( Figure 5C). All these data present a

PLZF inhibits the tumor growth of PC3 cells in xenograft nude mice
To further investigate the potential role of PLZF in vivo, PC3 cells stably transfected with PLZF-expressing vector or the empty vector were subcutaneously injected into the 6-week-old male athymic nude mice. All of the mice developed the tumor after one week and the tumor volumes were measured every other day. Transplanted tumors originated from PLZF overexpression cells grew more slowly than tumors from empty-vector cells ( Figure 6A). Twenty-six days later, mice were sacrificed, and tumors were picked. Figure 6B showed the representative macroscopic images for the PLZF-expressing and empty vector-carrying PC3 cellstransplanted tumors. Immunohistochemical staining for Ki67, a widely used proliferation marker, showed that PLZF-expressing tumors exhibited fewer Ki67-positive cells than the empty vectorcarrying tumors ( Figure 6C/D). Based on these results, we conclude that PLZF plays a critical role in suppression the tumorigenesis of PCa.

Discussion
Many studies have demonstrated that PTEN plays an important role in the development of prostate cancer, whose genomic deletion is closely associated with poor outcome of localized and androgen-independent prostate cancer [15,16,17]. By FISH assay, it has been shown that homozygous deletion of PTEN is more frequently in high-grade (GS = 8,9) prostate cancer specimens than in median-(GS = 7) and low-grade (GS = 5,6) ones [18]. Besides this, several groups have provided evidences that PTEN loss along with other pathways, such as RAS/MAPK and JNK, contributes to the progression of prostate cancer, especially the metastatic process [5,6]. However, the downstream executor of PTEN remains largely unknown. Based on our preliminary study in PTEN conditional knock-out mice and previous studies [9,10], we hypothesized that PLZF might be the downstream modulator cooperating with PTEN in regulating the progression of prostate cancer.
To test this hypothesis, we first performed immunohistochemical staining of PTEN and PLZF in a cohort of human PCa samples. The results presented lower expressions of PTEN and PLZF in high-grade prostate cancer specimens (GS = 8,9) than in mediate-grade (GS = 7) and well-differentiated (GS = 5,6) specimens. Further statistical analysis revealed a positive correlation between PTEN and PLZF, indicating that PLZF expression is influenced by PTEN in the progression of prostate cancer. Next   (21353,21348). FOXO3a TM overexpression significantly enhanced luciferase expression driven by the promoter of PLZF gene. Deletion-based luciferase reporter assay revealed that FOXO3a binding sites are located in the region of FHRE1 and FHRE2 within PLZF promoter. Individual abrogation of binding site assay demonstrated that FHRE1 is essential for FOXO3a binding on the promoter of PLZF while FHRE2 mutation does not influence significantly the luciferase expression. Chromatin immunoprecipitation assay confirmed the binding of FOXO3a is particularly located in the region of FHRE1 and FHRE2 instead of FHRE6 and FHRE7. In addition, ectopically expressed FOXO3a TM protein significantly increased PLZF expression in both mRNA and protein level in PC3 cells, and silence of FOXO3a by specific shRNA reduced PLZF expression. Furthermore, a significant positive correlation is seen in a cohort of human PCa samples between FOXO3a and PLZF expressions. Thus, we confirmed that PLZF mediates the PTEN signaling through the AKT/FOXO3a cascade. Its function in tumor suppression was further verified by xenograft animal experiments using PC3-PLZF cell line.
The promyelocytic leukemia zinc finger protein (PLZF), firstly discovered as part of the fusion protein with retinoic acid receptora (RARa) in acute promyelocytic leukemia (APL) [27], belongs to the family of Kruppel-type zinc finger proteins and contains a conserved POZ protein-binding domain at its N-terminal. PLZF 2/2 mice are prone to develop defect patterning in limb and axial skeleton [28]. Besides this, PLZF 2/2 mice displayed the disruption of osteoblastic differentiation during the early stages of pluoripotent differentiation of human mesenchymal stem cells [29]. Recently, large studies support that PLZF plays a key role in maintaining the pool of spermatogonial stem cells [10,30]. All these data indicated that PLZF might have a significant influence on maintaining the dynamic balance of self-renewal and differentiation. Since cancer stem cells share some characteristics with embyonic stem cells/progenitors, it should be a potential field to explore the activity of PLZF in cancer stem cells.
In addition, other studies have provided evidence showing PLZF is involved in the regulation of cell cycle progression and growth suppression [21] by repressing the expression of cyclin A2, c-myc, and c-kit [22,23,24]. Deregulation of PLZF activity is The folds of the relative pSV40-Renilla-normalized FHREs-Luc activity against cells without FOXO3a TM were calculated, shown as the means with bar as SD from three independent experiments. The P values for comparison were gotten by Student's t-test. B. PC3 cells were transfected with pGL3-PLZF or its truncations or empty vector pGL3-basic as indicated in the presence of FOXO3a TM expressing vector for 24 hours. The folds of the relative pSV40-Renilla-normalized FHREs-Luc activity against cells with empty vector pGL3-basic were calculated, shown as the means with bar as SD from three independent experiments. The P values for comparison were gotten by Student's t-test. C. PC3 cells were cross-linked and lysed, followed by immunoprecipitating with anti-FOXO3a antibody or non-specific IgG. Precipitated DNAs and total lysis were amplified by PCR with primers for district between FHRE1 and FHRE2 or district between FHRE6 and FHRE7, as shown in the top panel. The ChIP assay was repeated individually for three times. doi:10.1371/journal.pone.0077922.g004 demonstrated to be linked with the progression of multiple solid tumors including prostate cancer [19,20,31]. It has been shown that PLZF can directly bind to the promoter region of miR-221/222. Loss of PLZF in advanced melanoma cells unblocked miR-221/222, which in turn enhanced cell proliferation and invasion [31]. Besides microRNA, it has been shown that the transcriptional repression by PLZF is dependent on histone acetyltransferase (HAT), p300, which specifically acetlated the lysine residues in its C-terminus of zinc finger motif [32]. PLZF acetylation appears to promote its binding to DNA, thus enhancing its transcriptional repressor activity in this context. Based on these data, we believe that the loss of function of PLZF is an intriguing event during the development of solid tumors. In future studies, we should pay more attentions on the interaction network between PLZF and other DNA/protein/micorRNA, which would help us understand the role of PLZF in tumor progression more thoroughly, and shed new light on providing more clinical strategies.
Based on our in vitro and in vivo results, we propose that PLZF functions as a down-stream mediator of PTEN signaling and we define a signaling axis of PTEN/AKT/FOXO3a/PLZF in PCa, depicted in Figure 7. Our study underscores the importance of PLZF in suppressing prostate tumorigenesis and indicates further investigations on prostate-specific PLZF knock-out mice. Our findings may contribute to expand our knowledge on the prostate cancer progression and establish therapeutic strategies in the management of prostate cancer.

Ethics statement
Written informed consent was obtained from each patient for his enrollment. The study was approved by the Institutional Review Board of Renji Hospital, Jiao-Tong University School of Medicine (Shanghai, China). All clinical investigation was conducted in accordance with the principles expressed in the Declaration of Helsinki.
Cell culture and reagents PC3 and LNCaP cells (obtained from ATCC) were cultured in DMEM medium (GIBCO, Life Technology, NY, USA) and RPMI 1640 (GIBCO, Life Technology, NY, USA), respectively, supplemented with 10% new-born calf serum (NBCS, GIBCO), 1% penicillin/streptomycin (GIBCO), and 1% non-essential amino acid (GIBCO) in a humidified incubator in an atmosphere of 5% CO 2 , 37uC. HEK293T cells were applied to produce high titer retrovirus and cultured in DMEM medium. The PI3K inhibitor LY294002 was purchased from Cell Signaling Technology (Hitchin, United Kingdom), dissolved in DMSO according to the instruction.

Virus infection
Retroviruses were prepared by transient co-transfection of these plasmids with helper plasmids into 293T cells. PC3 and LNCaP cells were infected at approximately 70% confluence in DMEM supplemented with 8 mg/ml of polybrene. Forty-eight hours later, the medium was changed to DMEM medium with 10% FBS and added 4 mg/ml puromycin for 3 days to screen stable cell lines for further assay. Viruses-expressing AKT or PTEN infected into PC3 cells were generated from HEK293T by co-transfected with gagpol/vsvg system.

Luciferase assay
For the luciferase assay, FOXO3a TM , along with the luciferase reporter plasmid driven by PLZF promoter fragments and pSV40-Renilla, as an internal control, were co-transfected into the PC3 cells seeded in a 12-well plate. Twenty-four hours later, cells were lysed and analyzed for the luciferase assay according to the manufacture's instruction (Promega).

Chromatin immunoprecipitation (ChIP)
PC3 cells were cross-linked with 1% formaldehyde at 37uC for 10 minutes, and cells were scraped down in 500 ml RIPA lysis buffer (Beyond, Shanghai, China) supplemented with PMSF (Beyond) and protease inhibitor cocktail (Sigma-Aldrich Corp, St. Louis, MO, USA). Then DNA of the cells was sonicated and sheared to small fragments of 500-1000 bp with Sonicator ultrasonic processor (Misonix, Farmingdale, NY). Subsequently, the supernatant of the sonicated cells was collected, diluted and precleared by protein A+G agarose (Calbiochem). Furthermore, anti-human FOXO3a antibody (Cell Signaling Technology, Hitchin) was added to the supernatant for immunoprecipitation with normal pre-immuned mouse IgG (Santa Cruz Biotechnology) as a normal control. After overnight incubation, the protein A+G agarose were added and incubated for additional 3 hours and then washed with low-salt, high-salt and LiCl buffers serially, then the immunoprecipitated DNA was retrieved by 5 M NaCl at 65uC for 4 hours and purified with a DNA purification kit (Qiagen). PCR for the FHRE in the promoter was performed with specific primers: 59-gcagactcctgctgggctga-39 (P1, forward) and 59-gctgaaggccaggcggtaga-39 (P2, reverse); 59tccccatgcactcctgtcctct-39 (P3, forward) and 59-tccgggcttcgcctgacatcta-39 (P4, reverse).

Immunofluorescence
For immunofluorescence, the transfected PC3 cells were fixed in 4% paraformaldehyde for 10 min and incubated with antibodies against phosho-FOXO3a (diluted 1:100, Cell Signaling Technology, Hitchin). After washing with 16PBS buffer, cells were incubated with Cy3-conjugated secondary antibodies together with DAPI to stain nuclei. The slides were mounted and viewed under a fluorescence microscope (Nikon ECLIPSE Ti-S).

Immunohistochemistry
All 50 prostate cancer specimens collected from April 2003 to January 2010 by the Department of Urology at Renji Hospital, Jiao-Tong University School of Medicine (Shanghai, China), histologically graded according to the Gleason method [33], were separated into low-, moderate-, and high-grade subgroups, defined as cases with a combined GS = 5,6, GS = 7 or GS = 8,9, respectively. Serial sections of these specimens from formalinfixed and paraffin-embedded human Prostate cancer tissues were deparaffinised in xylene and rehydrated in graded ethanol solutions. Sections were pretreated in heated sodium citrate buffer (0.01 M, pH 6.0) for 15 minutes in 100uC followed by blocking the endogenous peroxidase with 0.3% hydrogen peroxide, and being incubated with PBS (0.1 M, pH 7.0) containing 10% donkey serum (Jackson ImmunoResearch Europe Ltd, Newmarket, Suffold, UK). Following overnight incubation with the antibodies including PLZF (diluted at 1:100, R&D), PTEN (diluted at 1:200, Cell Signaling Technology), or FOXO3a antibody (diluted at 1:100, Cell Signaling Technology) at 4uC, the sections were stained with the corresponding HRP-conjugated secondary antibody (KIT-9719 and KIT-9707, Mai-xin Biotechnology, Fuzhou, China). The visualization was performed with a diaminobenzidine (DAB) reaction (Mai-xin), resulting in a brown staining of structures containing the epitope. Slides were counterstained with hematoxylin before permanent mounting and then evaluated under a light microscope (Carl Zeiss, Maple Grove, MN, USA). IHC staining scores of PLZF and PTEN were judged depending on the staining intensity, which was scored by using a 4-tier system as negative (score = 1), weak (score = 2), moderate (score = 3) and strong (score = 4) staining.

Animal experiments
PLZF-overexpressing and its vector-control PC3 cells (2610 6 ), harvested, washed and resuspended in 100 ml serum-free DMEM, were injected subcutaneously into the left flank of the six-week old athymic nude-mice (six mice for each group) obtained from SLAC Inc. (Shanghai, China). When the tumors were detectable one week after injection, they were measured every two days with venire calipers for three more weeks. The tumor volume were calculated with the following formula: 1/2(the shortest diameter) 2 6(the longest diameter). All mice were sacrificed and all visible tumors were excised and separated for immunohistochemistry staining. All of the animals care and experimental protocols was approved by the animal committee of Shanghai Jiao-Tong University School of medicine (Shanghai, China).

Statistical analysis
Pearson's chi-square test and Spearman-rank correlation were employed to evaluate the results of immunohistochemistry. All statistical analyses were conducted with SAS Enterprise software (SAS Institute, Cary, NC). The values were exhibited as mean 6 SD. The Student's t-test was used for statistical analysis between two groups. Significant level was set at P,0.05. Figure S1 The formation of tumor in PTEN knock-out mice. The representative images of the tumor formed in one flank of testis (arrow) in PTEN knock-out mice. (TIF)