Figure 1.
Prostate carcinogenesis is increased in SPDEF−/− mice.
Experimental TRAMP/SPDEF−/− and control TRAMP mice were sacrificed at 23 weeks of age. A. Deletion of SPDEF increased the weight and sizes of prostate glands. Mean weights and diameters of prostate glands (±SD) were calculated from 8–9 mouse prostates per group. B. Efficiency of SPDEF deletion is shown by qRT-PCR. Total prostate RNA was prepared from TRAMP/SPDEF−/− and TRAMP mice. β-actin mRNA was used for normalization. Data represent means ± SD of three independent determinations using prostate tissue from n = 5–10 mice in each group. C. Increased mRNA levels of Cdc25b, Cyclin B1, Cyclin A2, Plk-1, CKS1, Aurora B and Topo2-alpha were found in TRAMP/SPDEF−/− prostates by qRT-PCR. Data represent means ± SD of three independent determinations (n = 5–10 mice in each group). D. Increased cellular proliferation in TRAMP/SPDEF−/− prostates. Mouse prostate glands were harvested 23 weeks after birth and used for immunohistochemistry with Ki-67 and PH3 antibodies. Number of positive cells were counted in 5 random microscope fields (n = 6 mice per group). Data represent mean ± SD. A p value<0.01 is shown with (**) and p value<0.05 is shown with (*). Magnification: panels D, 200×.
Figure 2.
Expression of SPDEF in TRAMP C2 prostate adenocarcinoma cells decreased cell growth, reduced migration and colony formation on soft agar.
A. qRT-PCR shows that SPDEF mRNA is increased in SPDEF OE cells (left panel). Overexpression of SPDEF reduced mRNAs of cell cycle regulatory genes. β-actin mRNA was used for normalization. B. Overexpression of SPDEF decreased proliferation of TRAMP C2 adenocarcinoma cells in vitro. Control and SPDEF-expressed TRAMP C2 cells were seeded in triplicates and counted at different time points using hemocytometer. C. Overexpression of SPDEF decreased migration of TRAMP C2 adenocarcinoma cells in vitro. Wound healing assay was used to measure cell migration. D. Increased expression of SPDEF decreased colony formation of TRAMP C2 cells on soft agar. The number of colonies were counted in 5 random fields in each of 3 individual wells per group. Data represent mean ± SD of three independent experiments. A p value<0.01 is shown with (**) and p value<0.05 is shown with (*).
Figure 3.
Expression of SPDEF in prostate adenocarcinoma cells decreased prostate carcinogenesis in orthotopic model.
Mouse prostates were harvested 5 weeks after inoculation of control TRAMP C2 adenocarcinoma cells or TRAMP C2 cells with stable expression of SPDEF (SPDEF OE). A. SPDEF decreased the growth of prostate tumors in orthotopic model. Mean weights of prostate glands (±SD) are shown (n = 10 for control TRAMP C2 cells, n = 5 for SPDEF OE cells). B. mRNA levels of SPDEF in tumors are shown by qRT-PCR. C. SPDEF decreased cellular proliferation as demonstrated by reduced numbers of Ki-67-positive and PH3-positive cells. Magnification is ×100. D. Percentage of Ki-67-positive and PH3-positive cells were counted in five random microscope fields (n = 3 mice per group, left panels). Decreased mRNA levels of proliferation-specific genes in SPDEF OE prostates were found by qRT-PCR (right panel). Data represent means ± SD of three independent determinations using prostate tissue from n = 5–10 mice in each group. A p value<0.01 is shown with (**) and p value<0.05 is shown with (*).
Figure 4.
Transgenic expression of SPDEF in prostate epithelium decreased prostate carcinogenesis.
A. Schematic drawing shows Dox-inducible expression of SPDEF and TRAMP transgenes in prostate epithelial cells. B. Decreased prostate carcinogenesis in TRAMP/SPDEF OE mice. Experimental TRAMP/SPDEF OE and control TRAMP mice were given Dox at 4 weeks of age and sacrificed at 25 weeks of age. Mean weight of prostate glands (±SD) was calculated from 7–10 mouse prostates per group. A p value<0.01 is shown with asterisk (**). C. TRAMP/SPDEF OE prostates show increased SPDEF mRNA (left panel) and decreased mRNA levels of cell cycle regulatory genes (right panel). D. Decreased number of proliferating cells in TRAMP/SPDEF OE prostates. Prostate sections were stained with PH3 antibody. The number of PH3-positive cells was counted using 5 random fields in each of 3 individual mice per group. Data represent mean ± SD. A p value<0.01 is shown with (**) and p value<0.05 is shown with (*). Magnification: panels D, 200×.
Figure 5.
SPDEF and Foxm1 are inversely correlated in prostate carcinogenesis.
A. In transgenic TRAMP/SPDEF OE mice, over-expression of SPDEF in prostate epithelial cells decreased Foxm1 mRNA (left panel) and protein (right panels) in prostate tumors. Experimental TRAMP/SPDEF OE and control TRAMP mice were sacrificed at 25 weeks of age. B. In orthotopic mouse model, SPDEF inhibited Foxm1 mRNA and protein levels during prostate carcinogenesis. Lentiviral expression of SPDEF in TRAMP C2 prostate adenocarcinoma cells decreased Foxm1 mRNA shown by qRT-PCR. β-actin mRNA was used for normalization. The decrease of Foxm1 staining in prostate tumors is shown by immunohistochemistry. Mouse prostates were harvested 5 weeks after inoculation of either control TRAMP C2 cells or TRAMP C2 cells expressing SPDEF (SPDEF OE). C. In transgenic TRAMP/SPDEF−/− mice, depletion of SPDEF increased Foxm1 mRNA (left panel) and protein levels (right panels). Experimental TRAMP/SPDEF−/− and control TRAMP mice were sacrificed at 23 weeks of age. Data represent means ± SD of three independent determinations (n = 3–5 mice in each group). Magnification: 200×. A p value<0.01 is shown with (**) and p value<0.05 is shown with (*). C. Magnification: 200×. A p value<0.01 is shown with (*).
Figure 6.
SPDEF expression is inversely correlated with Foxm1 expression in human prostate cancer and of prognostic value for the prostate carcinoma patient survival.
The data from two human prostate cancer microarray datasets, GSE21034 [30] and GSE16560 [31] were downloaded from the GEO archive. Expression levels were compared between Indolent and Lethal prostate cancers (A), and High-risk and Low-risk sample groups (B). High-risk samples were derived from patients with surviving less than 12 months and low-risk samples from patients surviving more than 192 months. C. Two-gene expression signature predicts poor patient survival. Kaplan-Meier survival analysis of prostate cancer patients using dataset GSE16560 [31]. Patients were stratified by the expression level of FOXM1 or SPEDF, or both together. The group with “high” FOXM1 and “low” SPDEF expression had the worst outcome (median survival time of 55.5 months).
Figure 7.
Re-expression of Foxm1 in the SPDEF-positive prostate adenocarcinoma cells restored tumor cell proliferation in vitro and in vivo.
We used SPDEF-overexpressing prostate adenocarcinoma TRAMP C2 cells (SPDEF OE cells) to stably express Foxm1 (SPDEF/Foxm1 OE cells). A. Foxm1 mRNA levels were determined by qRT-PCR. Growth curves demonstrated that re-expression of Foxm1 in SPDEF OE cells restored the growth of these cells in culture. B. Flow cytometery shows that the re-expression of Foxm1 in SPDEF OE cells increased entry of synchronized cells into S phase at 12 hours after serum addition in vitro. C. In orthotopic mouse model of prostate cancer, re-expression of Foxm1 in SPDEF-overexpressing cancer cells restored tumor sizes that have been decreased after SPDEF expression (left panel). Foxm1 and SPDEF mRNAs in tumor tissues are shown by qRT-PCR (middle panels). Protein levels of Foxm1 and SPDEF are shown by Western blot (right panel). D. Re-expression of Foxm1 restored cellular proliferation in SPDEF-overexpressing prostate tumor cells as demonstrated by increased numbers of Ki-67-positive (upper panels) and PH3-positive (bottom panels) cells. Percentages of Ki-67-positive (upper panels) and PH3-positive cells (bottom panels) were counted in five random microscope fields (n = 3 mice per group, right panels). Magnification is ×100. A p value<0.01 is shown with (**) and p value<0.05 is shown with (*).
Figure 8.
SPDEF represses the Foxm1 promoter.
A. Schematic drawing of the mouse Foxm1 promoter shows the presence of an evolutionary conserved Foxm1 binding site (black oval) and three SPDEF binding sites (white boxes). B. Schematically shown the luciferase (Luc) reporter constructs: Luc I, includes the −3.7 Kb Foxm1 promoter region; Luc II-IV, include one of its deletion mutants; Luc V, includes a construct with mutations in Foxm1 site; Luc VI, includes a construct with mutations in SPDEF site. TRAMP C2 cells were transfected with CMV-Foxm1b or CMV-SPDEF expression vectors and one of the Foxm1 promoter LUC plasmids. CMV-empty plasmid was used as a control. Dual LUC assays were used to determine LUC activity. Transcriptional induction is shown as a fold change relative to CMV-empty vector (± SD). A p value<0.01 is shown with (**) and p value<0.05 is shown with (*). C. Western blot shows efficient expression of SPDEF in TRAMP C2 cells after lentiviral transduction (left panel). ChIP assay was performed in control TRAMP C2 cells and TRAMP C2 cells overexpressing SPDEF (SPDEF OE). In control cell, Foxm1 is bound to its own −745/−660 bp promoter region (Foxm1 IP). In SPDEF OE cells, SPDEF is bound to the −745/−660 bp Foxm1 promoter region (SPDEF IP) and the binding of Foxm1 to this region is lost. Neither Foxm1, nor SPDEF bound to the −3663/−3653 and to the −1067/−1057 bp Foxm1 sites (grey arrows in A). D. Schematic drawing shows that SPDEF protein physically binds to the −745/−660 bp Foxm1 promoter region and interferes with Foxm1 binding to the same region.