PLZF mediates the PTEN/AKT/FOXO3a signaling in suppression of prostate tumorigenesis

Abstract

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.

Figure 1. Correlation of PTEN and PLZF expressions in prostate cancer tissues.

A. The representative immunohistochemical staining for PTEN and PLZF of human prostate cancer specimens of low (GS = 5, 6), moderate (GS = 7) and high grades (GS = 8, 9) by the Gleason score. Scale bar, 50 µm. B, C. The mean staining scores of PTEN (B) and PLZF (C) were evaluated in low, moderate and high grades of prostate cancer specimens. Data were shown as mean ± SD. The P values for comparison between line-linked groups were gotten by Student's t-test. D. Box plot of PLZF staining score in tumors with low and high PTEN expression. Horizontal lines represent the median; the bottom and top of the boxes represent the 25th and 75th percentiles, respectively; and the vertical bars represent the range of data. Data were analyzed by Pearson's chi-square test. E. The percentage of tumors with high and low PLZF expression in the PTEN^high and PTEN^low groups of subjects. Data were analyzed by Spearman-rank correlation.

Figure 2. PLZF expression is influenced by PTEN in Prostate cancer cells.

A. Western blotting showing that rescued expression of PTEN and the resultant induction of PLZF expression in PC3 and LNCaP cells. PC3 and LNCaP cells were transfected with PTEN or its vehicle plasmid, followed by Western blots for the proteins as indicated. B. PC3 cells were treated with LY294002 for the indicated times, and AKT and its phosphorylated form (p-AKT) were detected by western blots with β-actin as loading control. C. PC3 cells were treated with LY294002 for the indicated time. Then, PLZF mRNA was detected by Q-PCR. Folds of PLZF mRNA against cells in LY294002-treated for 0 hour were calculated, and results were shown as means with bar as SD in three independent experiments with triplicate each. The P values for comparison between line-linked groups were gotten by Student's t-test. D. PC3 cells and LNCaP cells were treated with LY294002 for the indicated time. PLZF proteins were detected by Western blots with β-actin as loading control. All experiments were repeated individually for three times.

Figure 3. FOXO3a regulates PLZF expression in prostate cancer cells.

A. Wild-type AKT, AKT^CA, and AKT^DN were transfected into PC3 cells, and Western blots were performed with the proteins as indicated. B. AKT^CA and AKT^DN were transfected into PC3 cells, followed by immunofluorescent staining of AKT and FOXO3a with re-stain of DAPI. The representative images were recorded under a 40× objective. C. Flag-tagged FOXO3a and FOXO3a™ were transfected into PC3 cells, followed by immunofluorescent staining of Flag and PLZF with re-stain of DAPI. The representative images were recorded under a 40× objective. D, E. Wild-type FOXO3a and triple mutant, FOXO3a™ were transfected into PC3 cells respectively, and Q-PCR (D) and Western blots (E) were performed as indicated. In panel D, folds of PLZF mRNA against cells transfected with vehicle were calculated, and results were shown as means with bar as SD in three independent experiments with triplicate each. The P values for comparison between line-linked groups were gotten by Student's t-test. F. ShRNAs specifically against FOXO3a along with its non-specific shRNA (NC) were respectively transfected into LNCaP cells, and Western blots were performed with FOXO3a and PLZF with β-actin as loading control. All the experiments were repeated individually for three times.

Figure 4. PLZF is a direct target gene of FOXO3a.

A. The luciferase reporter plasmid pGL3-PLZF, shown in the top and middle panel, is driven by seven putative FHREs in PLZF promoter. In the top diagram, black ovals and the empty circle represent the putative FHREs and the transcriptional start site, respectively. PC3 cells were transfected with 100 ng of pGL3-PLZF and/or 200 ng of FOXO3a™ together with pSV40-renilla for 24 hours. The folds of the relative pSV40-Renilla-normalized FHREs-Luc activity against cells without FOXO3a™ 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™ 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.

Figure 5. FOXO3a expressions in prostate cancer tissues.

A. The representative immunohistochemical staining for FOXO3a of human prostate cancer specimens of low (GS = 5, 6), moderate (GS = 7) and high grades (GS = 8, 9) by the Gleason score. Scale bar, 50 µm. B. The mean staining scores of FOXO3a was evaluated in low, moderate and high grades of prostate cancer specimens. Data were shown as mean ± SD. The P values for comparison between line-linked groups were gotten by Student's t-test. C. The Spearman correlation of FOXO3a and PLZF expression to each other was illustrated.

Figure 6. PLZF inhibits the tumor growth in nude mice.

A. The tumor volumes (six mice for each group) were measured at the indicated times, shown as the means with bars as SD. The P values for comparison were gotten by Student's t-test. The inserted western blots showed PLZF expression. B. Representative macroscopic images for tumors 26 days after subcutaneous injection. C. Representative immunohistochemistry staining of Ki67 (scale bar, 50 µm) for PLZF-expressing and empty vector-carrying tumors. The representative images were recorded under a 40× objective. D. The statistical analysis of the accounts of the Ki67 positive cells was displayed on the right panel, shown as the means with bar as SD. The P values for comparison were gotten by Student's t-test.

Figure 7. The putative PTEN/FOXO3a/PLZF signaling pathway in the development of prostate cancer.