Androgen-responsive serum response factor target genes regulate prostate cancer cell migration

Abstract

Progression of prostate cancer (CaP) relies on androgen receptor (AR) signaling, but AR-dependent events that underlie the lethal phenotype remain unknown. Recently, an indirect mechanism of androgen action in which effects of AR on CaP cells are mediated by Serum Response Factor (SRF) has been identified. This is the first mode of androgen action to be associated with aggressive CaP and disease recurrence. The manner in which androgen-responsive SRF activity controls aggressive CaP cell behavior is unknown. Here, the contribution of two representative SRF effector genes that are underexpressed, calponin 2 (CNN2), or overexpressed, sidekick-homolog 1 (SDK1), in clinical CaP specimens is studied. AR- and SRF- dependency of CNN2 and SDK1 expression was verified using synthetic and natural androgens, antiandrogens, and small interfering RNAs targeting AR or SRF, and evaluating the kinetics of androgen induction and SRF binding to endogenously and exogenously expressed regulatory gene regions in AR-positive CaP model systems that mimic the transition from androgen-stimulated to castration-recurrent disease. Small interfering RNA-mediated deregulation of CNN2 or SDK1 expression did not affect CaP cell proliferation or apoptosis but had marked effects on CaP cell morphology and actin cytoskeleton organization. Loss of CNN2 induced cellular protrusions and increased CaP cell migration, whereas silencing of SDK1 led to cell rounding and blunted CaP cell migration. Changes in cell migration did not involve epithelial-mesenchymal transition but correlated with altered β1-integrin expression. Taken together, individual androgen-responsive SRF target genes affect CaP cell behavior by modulating cell migration, which may have implications for therapeutic intervention downstream of AR and SRF.

Fig. 3.

CNN2 and SDK1 regulate CaP cell morphology, organization of the actin cytoskeleton and migration. (A) LNCaP cells were transfected with non-targeting control siRNA (top row), or siRNA targeting the expression of CNN2 (middle row) or SDK1 (bottom row). After 48 or 96 h, cells were fixed and stained with rhodamine phalloidin (red) and counterstained with DAPI (blue). (B) Quantification of morphological CaP cell features following siRNA-mediated silencing of CNN2 or SDK1. Each slide was divided into four quadrants and 12 or 13 random cells per quadrant (i.e. 50 cells in total) were evaluated for cell features such as the number of nuclei, presence of cell extensions, number of cell extensions, presence of stress fibers and number of fibers that are present. A more detailed description of individual cell data is provided in Supplementary Tables 2–4, available at Carcinogenesis Online. Student’s t-tests were done to determine statistical significance between control and CNN2 siRNA groups and between control and SDK1 siRNA groups. (C) LNCaP cells were transfected with non-targeting control siRNA, or siRNAs directed against CNN2 or SDK1 expression. The next day, cells were seeded into culture inserts. One day later, inserts were removed (day 0). Cell migration was evaluated at days 1, 2, and 3. (D) Quantification of wound closure. Black columns, cells transfected with non-targeting control siRNA; gray columns, cells transfected with siRNA targeting CNN2 (top panel) or SDK1 (bottom panel). Columns, means of values from five measurements; bars, standard error of the mean. Cell proliferation was assessed in parallel using MTS assays at day 1 as before (right panels).

Fig. 1.

Expression of CNN2 and SDK1 is androgen regulated in an SRF-dependent manner. (A) LNCaP cells were treated with 5nM of the synthetic androgen R1881 (+) or ethanol vehicle (−) for 48 h. Cells were harvested and protein and RNA were isolated. RNA was converted into cDNA and real-time RT–PCR was done using primers that target CNN2 and SDK1 expression. Expression values were normalized to glyceraldehyde 3-phosphate dehydrogenase expression levels and expressed as relative expression, where the value obtained from one of the vehicle-treated samples was taken as 1. Black columns, vehicle-treated samples; gray columns, R1881-treated samples. Columns, means of values obtained from three independent biological replicates; bars, standard error of the mean values (top panel). Western blotting was performed using antibodies directed against CNN2 and SDK1. Blots were reprobed for β-actin (β-act) to control for loading differences (bottom panel). (B) LNCaP cells were transfected with siRNA directed against AR or a non-targeting control (c) siRNA. One day later, medium was replaced by CSS-supplemented medium. The next day, medium was replaced and cells were treated with vehicle or R1881 (5nM). After 48 h, cells were harvested for RNA and protein isolation. Real-time RT–PCR using primers directed against CNN2 and SDK1 was performed as described (left panel). Western blotting was performed with antibodies directed against CNN2, SDK1, AR and β-actin (β-act) (right panel). (C) LNCaP cells were treated with ethanol vehicle, R1881 (1nM), Casodex (10 μM, csdx) or R1881 and an excess of Casodex for 48 h. Cells were harvested for RNA, and cDNA synthesis followed by real-time RT–PCR was performed (left panel). LNCaP cells were treated with R1881 or Casodex at various concentrations either alone or in combination, and whole-cell protein extracts were subjected to western blotting using antibodies directed against CNN2 or SDK1 as above (right panel). (D) LNCaP cells were treated with 5nM R1881 or vehicle and harvested for RNA isolation 4 or 16 h later. Real-time RT–PCR was performed using primers directed against CNN2, SDK1 orprostate-specific antigen as above (top panel). LNCaP cells were treated with vehicle or R1881 (5 nM) and harvested after 0, 4, 8, 16, 24, 48 and 72 h. Western blot analysis was performed using antibodies directed against CNN2, SDK1 and β-actin (β-act) (bottom panel). (E) LNCaP cells were transfected with non-targeting siRNAs or siRNAs directed against SRF. One day later, medium was replaced by CSS-supplemented medium. The next day, medium was replaced and cells were treated with vehicle or R1881 (5nM). After 48 h, cells were harvested for protein and RNA isolation, and western blotting (right panel) and real-time RT–PCR (left panel) were done to evaluate CNN2 and SDK1 expression levels. (F) LNCaP cells were treated with either vehicle or R1881 (5nM) for 16 h. ChIP assays were performed as before using an antibody directed against SRF or non-targeting IgG. CNN2, CArG box containing CNN2 promoter fragment; control, similarly sized non-CArG box-containing exonic CNN2 gene fragment. (G) Graphical representation of the structure of the CNN2 promoter-reporter constructs. CNN2wt-luc, wild-type construct containing a 982 bp CNN2 promoter fragment that harbors a CArG box 133 base pairs upstream of the transcriptional start site; CNN2mut-luc, mutant construct in which the CArG box has been mutated. (H) LNCaP cells were transfected with CNN2wt-luc or CNN2mut-luc and treated with vehicle, R1881 and/or Casodex (csdx) for 48 h (left panel and right panel), with vehicle or R1881 in the presence of non-targeting siRNA or SRF-directed siRNA (middle panel). The next day, cells were treated with ethanol vehicle or R1881 (5nM). After 48 h, a luciferase assay was done. Columns, means of values obtained from three independent biological replicates; bars, standard error of the mean.

Fig. 2.

CNN2 and SDK1 do not regulate CaP cell proliferation or apoptosis. (A) LNCaP cells were transfected with non-targeting control (c) siRNA (black columns) or siRNA directed against CNN2 and SDK1 (gray columns). Cell proliferation was assessed by performing MTS assays 48 h (top panels) and 96 h (bottom panels) after transfection under regular FBS-supplemented culture conditions. Absorbance at 490 nm (A490) was read. Columns, means of values from five biological replicates; bars, standard error of the mean values. (B) LNCaP cells were transfected with non-targeting control siRNAs (c) or siRNAs directed against CNN2 and SDK1. Ninety-six hours after transfection, cells were harvested for protein isolation and western blotting was performed using antibodies against PARP and β-actin (β-act) (left panel). Protein extracts from LNCaP cells that were treated with staurosporine (stauro) were subjected to western blotting using antibodies directed against PARP and β-actin (β-act) (right panel).

Fig. 4.

Effect of CNN2 and SDK1 silencing on expression of EMT markers and β1-integrin. LNCaP cells were seeded and transfected with non-targeting control (c) siRNAs, or siRNAs directed against CNN2 or SDK1 expression. Four days after transfection, cells were harvested and western blot analysis was performed using antibodies directed against the EMT markers β-catenin, claudin-1, E-cadherin and ZO-1 (A), β1-integrin (B) and β-actin (β-act).

Fig. 5.

Expression of CNN2 and SDK1 is androgen regulated in an SRF-dependent manner in CR-CaP cell lines. (A) C4-2 cells were treated with vehicle or R1881 (5nM) for 96 h. Cells were harvested, total protein was isolated and subjected to western blotting using antibodies directed against CNN2, SDK1 and β-actin (β-act) (top panel). C4-2 cells were transfected with CNN2wt-luc and treated with vehicle, R1881 (1nM) and/or an excess of Casodex (csdx) (bottom panel, left), or with vehicle or R1881 (5nM) in the presence of non-targeting control siRNA or siRNA targeting SRF (bottom panel, right). After 2 days, cells were harvested and a luciferase activity was done. Columns, means of values obtained from three independent biological replicates; bars, standard error of the mean. Inset, western blotting control for the efficiency of siRNA-mediated SRF silencing. (B) The experiments described under (A) were performed using LN-Rf cells.

Fig. 6.

CNN2 and SDK1 regulate organization of the actin cytoskeleton and cell migration in CR-CaP cells. C4-2 cells (A) and LN-Rf (B) cells were seeded on cover slips and transfected with non-targeting control siRNA (left image), or siRNAs targeted against CNN2 (middle image) and SDK1 (right image). After 96 h, cells were fixed and stained with rhodamine phalloidin as described (left panel). Protein extracts from C4-2 and LN-Rf cells that were transfected as above were analyzed via western blotting using antibodies directed against CNN2, SDK1 or β-actin (β-act) (right panel).