In total, 133 patients who underwent radical prostatectomy at Yamaguchi University Hospital from November 2000 to September 2016 were enrolled in the present study. All patients were diagnosed pathologically with prostate cancer. Detailed patient characteristics are presented in Table I. The present study was approved by the Institutional Ethics Committee of the Graduate School of Medicine of Yamaguchi University and written informed consent was obtained from all individuals enrolled in the study.

Formalin-fixed and paraffin-embedded tissue specimens were subjected to H&E staining and immunohistochemical (IHC) staining. For each sample, 3-µm-thick sections were deparaffinized in xylene, dehydrated in ethanol, and incubated in a 0.3% hydrogen peroxide solution in methanol for 10 min at room temperature. The sections were then microwaved in a 0.01 M citrate-buffered solution (pH 6.0) for 15 min and covered in blocking solution (IMMUNO SHOT; Cosmo Bio Co., Ltd.) for 30 min at room temperature. Then, a primary antibody [anti-ARHGAP29 (1:200 dilution; cat. no. HPA026534; Atlas Antibodies) or anti-YAP (1:200 dilution; product no. 14074; Cell Signaling Technology, Inc.)] was incubated according to the manufacturers' instructions overnight at 4°C, followed by incubation with the respective secondary antibody (N-Histofine Simple Stain MAX PO MULTI; cat. no. 414152F; Nichirei Biosciences, Inc.) for 30 min at room temperature. To evaluate IHC staining, the H-score was used in the present study. Briefly, >500 tumor cells were counted in five different fields of vision in each section (×100, magnification), and the H-score was calculated by multiplying the percentage of positive cells by the intensity (strongly stained, 3×; moderately stained, 2×; weakly stained, 1×), yielding a possible range of 0–300 (20–22). Two independent examiners (KS and HM) judged the scores and the mean score was set to the representative score. Cut-off of the H-score was determined by receiver operating characteristic (ROC) curve.

Four primary prostate cancer cell lines (22Rv1, ATCC no. CRL-2505; LNCaP, ATCC no. CRL-1740; DU145, ATCC no. HTB-81; PC-3, ATCC no. CRL-1435) were purchased from the American Type Culture Collection. Cells were cultured in RPMI-1640 and DMEM (Life Technologies; Thermo Fisher Scientific, Inc.) supplemented with 10% fetal bovine serum (Biological Industries) and maintained in humidified incubators with 5% CO₂ at 37°C.

si-ARHGAP29 and control siRNAs were obtained from Life Technologies; Thermo Fisher Scientific, Inc. siRNA sequences were as follows: ARHGAP29-#1 siRNA sense, 5′-GCAUAGGUGUUGUUGAUCAtt-3′ and antisense, 5′-UGAUCAACAACACCUAUGCta-3′; ARHGAP29-#2 siRNA sense, 5′-GACCAAGGCUAAAACGAAUtt-3′ and antisense, 5′-AUUCGUUUUAGCCUUGGUCtc-3′. The PC-3 cell line was transiently transfected with siRNA using Lipofectamine RNAi MAX (Thermo Fisher Scientific, Inc.), according to the manufacturer's instructions. After transfection, cells were incubated at 37°C in a CO₂ incubator for 48 h. Quantitative evaluations of mRNA and protein expression were performed by western blotting and RT-qPCR, respectively.

A mammalian expression of HA tagged ARHGAP29 (#104154) was purchased from Addgene, Inc. Cells were seeded on culture dishes at density of 1×10⁵/well in a 6-well plate, and the pcDNA3.1 empty vector plasmid (mock) (Thermo Fisher Scientific, Inc.) or 2 µg of ARHGAP29 expressing plasmid were transfected using X-tremeGENE HP DNA transfection Reagent (Sigma-Aldrich; Merck KGaA) for 48 h, according to the manufacturer's instructions.

Regarding the DU145 cell line, plasmid transfection was performed via electroporesis system using an Amaxa cell line Nucleofector Kit L (cat. no. VACA-1005; Lonza Group, Ltd.) according to the manufacturer's instructions. Prior to electroporation, 1×10⁶ DU145 cells were centrifuged at 90 × g for 5 min, resuspended in 100 µl of Nucleofector solution and mixed with 2 µg of pmaxGFP or 2 µg of ARHGAP29 plasmid. The aforementioned cells were transferred to cuvettes and immediately electroporated based on the DU145 program (Nucleofector Program A-023) using Nucleofector 2b Device. After electroporation, cells were incubated in the cuvette at room temperature for 10 min and then 500 µl of pre-warmed RPMI-1640 medium supplemented with 10% FBS were added to the cuvette. Cells were transferred to a 6-well plate and incubated at 37°C 5% CO₂ overnight. The day after electroporation, cells were centrifuged and the medium was replaced by RPMI-1640 supplemented with 10% FBS and incubated for 48 h and then performed subsequent experiments were performed.

We created cDNA by reverse transcription of mRNAs extracted from each prostate cancer cell line (22Rv1, LNCaP, DU145 and PC-3), using iScript Advanced cDNA Synthesis Kit for RT-qPCR (cat. no. 1725037; Bio-Rad Laboratories, Inc.). Quantitative real-time RT-PCR was performed in triplicate with an Applied Biosystems StepOnePlus using TaqMan universal PCR master mix (Applied Biosystems; Thermo Fisher Scientific, Inc.), according to the manufacturer's protocols. The TaqMan probes and primers were purchased from Applied Biosystems. Human GAPDH (assay ID: 02786624) was used as an endogenous control. Levels of ARHGAP29 (assay ID: 00191351) and MMP-2 (assay ID: 01548727) RNA expression were determined using StepOnePlus software (version 2.2.2; Applied Biosystems; Thermo Fisher Scientific, Inc.). The miRNA expression levels were determined using the 2^−ΔΔCq method (23). The cycling conditions consisted of an initial denaturation at 95°C for 30 sec and PCR at 40 cycles at 95°C for 5 sec and 60°C for 30 sec.

Total RNA was isolated from cells (PC-3-si-NC and si-ARHGAP29), and an RT² Profiler PCR Array (Qiagen RT² Profiler PCR Array Human Cell Motility; cat. no. PAHS-128Z, product no. 330231) was used to examine the expression patterns of genes involved in human cell motility, according to the manufacturer's instructions. We analyzed the gene expression levels and produced a heatmap using the web-based software ‘RT2 Profiler PCR Array’ Data Analysis version 3.5 (Qiagen, Inc.).

Cells samples were lysed in RIPA buffer (cat. no. 89900; Thermo Fisher Scientific, Inc.) supplemented with 1% protease inhibitors (cOmplete™, Mini, cat. no. 04693124001; Sigma-Aldrich) for total protein extraction. We quantified the concentration of total proteins using BCA. Each lysate sample (30 µg protein) was separated by 4–20% SDS-PAGE (Mini-PROTEAN TGX Stain-Free Gels, cat. no. 4568095; Bio-Rad Laboratories), and then electro-transferred to a PVDF membrane. After blocking in 5% dry non-fat milk or 5% BSA for 1 h at room temperature, the membranes were incubated with a primary antibody overnight at 4°C. After washing in TBS with 0.05% Tween-20 (TBST), the membranes were incubated with an HRP-conjugated secondary antibody for 1 h at room temperature. After washing with TBST, signals were detected using an ECL detection system (ChemiDoc™ XRS+; Bio-Rad Laboratories, Inc). Primary antibodies were as follows: anti-ARHGAP29 (product code ab85853, 1:2,000 dilution), anti-AR (product code ab133273, 1:1,000 dilution), anti-F-actin (product code ab205, 1:500 dilution) and anti-MMP-2 (product code ab97779, 1:1,000 dilution) from Abcam and anti-YAP (cat. no. 14074S; 1:1,000 dilution), anti-phosphorylated YAP (cat. no. 13008S; 1:1,000 dilution), anti-GAPDH (cat. no. 5174S; 1:1,000 dilution), anti-Cofilin (cat. no. 5175T; 1:1,000 dilution) and anti-phospho-Cofilin (cat. no. 3313T; 1:1,000 dilution) were obtained from Cell Signaling Technology Inc. Secondary antibodies were as follows: goat anti-rabbit IgG H&L (HRP) (product code ab6721; 1:10,000 dilution) and goat anti-mouse IgG H&L (HRP) (product code ab6789; 1:10,000 dilution) from Abcam. GAPDH was used for protein normalization. We performed densitometry using the public domain free software ImageJ (version 1.51; National Institutes of Health).

Cell viability was assessed using an MTS assay (CellTiter 96 AQ_ueous One Solution Cell Proliferation Assay; Promega Corporation). After the cells were seeded at density of 5×10³/well in a 96-well plate, cell viability was measured at 24, 48, and 72 h at an OD of 490 nm. Data are expressed as the mean ± SD of three independent experiments. Cell invasion assays were performed using a CytoSelect 24-well cell invasion assay kit (Cell BioLabs, Inc.). The CytoSelect™ Cell Invasion Assay Kit contains polycarbonate membrane inserts (8-µm pore size) in a 24-well plate. The upper surface of the insert membrane is coated with a uniform layer of dried basement membrane matrix solution. This basement membrane layer serves as a barrier to discriminate invasive cells from non-invasive cells. A cell suspension containing 0.5–1.0×10⁶ cells/ml was placed in upper chamber in serum-free media. A total of 500 µl of media containing 10% fetal bovine serum was added to the lower well of the invasion plate. After 48 h of incubation at 37°C with 5% CO₂, cells invaded through the basement membrane layer and clung to the bottom of the insert membrane. Non-invasive cells remained in the upper chamber. After removal of non-invasive cells, invasive cells were stained for 10 min at room temperature using Cell Stain Solution (Part no. 11002; CytoSelect 24-well cell invasion assay kit) and then quantified. Each insert was transferred to an empty well, 200 µl of Extraction Solution (Part no. 11003; CytoSelect 24-well cell invasion assay kit) was added per well and then incubation followed for 10 min on an orbital shaker. Subsequently 100 µl from each sample was transferred to a 96-well microtiter plate and the OD 560 nm of each sample was measured on a plate reader, according to the manufacturer's instructions.

The Cancer Genome Atlas (TCGA) accessed from the data portal of the National Cancer Institute Home Page (http://cancergenome.nih.gov/) was used for comparison with our data.

Categorical variables were compared by the Chi-squared test. Continuous variables were analyzed using the unpaired Student's t-test when comparing two groups. One-way ANOVA followed by Tukey-Kramer test were used when comparing more than two groups. Survival analysis was estimated by the Kaplan-Meier method and compared by the log-rank test. A Cox proportional hazards regression model was used in the multivariable analysis to identify risk factors for disease progression. Statistical analysis was performed using JMP software (Pro.13; SAS Institute). P-values were two-sided, and statistical significance was defined as P<0.05 in all tests. Regarding protein expression, bivariate analysis was performed and a ROC curve was constructed using JMP software to set the cutoff value and determine the high/low expression of proteins (24).

RT-qPCR and western blotting were performed to clarify whether there was a difference in the expression of ARHGAP29 between prostate cancer cell lines depending on AR and YAP expression. AR was expressed in 22Rv1 and LNCaP cell lines. YAP was expressed in all four prostate cancer cell lines (Fig. 1A and B). The expression level of YAP was higher in DU-145 and PC-3 cells than in LNCaP and 22Rv1 cells. ARHGAP29 protein expression was higher in PC-3 cells than in the other cell lines. F-actin was the most weakly expressed in PC-3 cells compared with the other cell lines (Fig. 1C).

After downregulation (PC-3 cells) or upregulation (LNCaP and DU145 cells) of ARHGAP29 in prostate cancer cell lines (Fig. 2A and B), the expression of several proteins was examined. Based on a recent study (18), the RhoA-LIMK-cofilin signaling pathway has been revealed to be affected by ARHGAP29 in a gastric cancer cell line. Therefore, certain related genes (cofilin, p-cofilin and F-actin) were analyzed by western blotting (Fig. 2B).

After almost complete knockdown of ARHGAP in PC-3 cells, phosphorylated cofilin and F-actin were increased and cofilin expression was unchanged. In contrast, after overexpression of ARHGAP29 in DU145 cells, F-actin was slightly decreased but phosphorylated cofilin was not altered. YAP and phosphorylated YAP were slightly recovered without significant differences in si-ARHGAP29 PC-3 transfectants compared with the si-NC control. Conversely, YAP was decreased after overexpression of ARHGAP29 in DU145 cells. Expression of these proteins relative to that of the housekeeping gene GAPDH and the ratio of phosphorylated protein to non-phosphorylated protein (YAP and cofilin) are presented in Fig. 2C and D.

Functional analyses by MTS and cell invasion assays were also performed in these three cell lines (Fig. 3). Cell viability and invasion were significantly decreased after downregulation of ARHGAP29 in PC-3 cells (Fig. 3A and B). After upregulation of ARHGAP29 in LNCaP and DU145 cells, cell viability and invasion were significantly increased (Fig. 3A and B).

Based on the functional analyses, ARHGAP29 may be involved in cell proliferation or invasion. To determine new therapeutic targets or genes related to ARHGAP29 in prostate cancer cells, Qiagen RT² Profiler PCR Array Human Cell Motility was used.

The pre-designed array included 84 genes related to cell motility (Fig. 4A). Data analysis was performed using the web-based software ‘RT2 Profiler PCR Array’ Data Analysis version 3.5 as aforementioned. A heatmap is presented in Fig. 4B. When the boundary was set at 3, one gene (STAT3) was upregulated and numerous genes (including CSF1, ACTN3 and HGF) were downregulated after knocking down ARHGAP29 in PC3 cells (Fig. 4C). Regulation of some proteins, such as HGF, RHO, CAPN1, was validated by western blotting. However, there was no difference in the expression of these proteins between si-NC and si-ARHGAP29 cells (data not shown). Among the downregulated genes of the 84 genes in the array, active MMP2 expression was significantly decreased at mRNA and protein levels after knockdown of ARHGAP29 in PC3 cells (Figs. 2B and C and 4D).

The expression level of ARHGAP29 was evaluated by IHC in 133 prostate cancer patients who had undergone radical prostatectomy. Representative images of YAP and ARHGAP29 staining in prostate cancer specimens (negative and positive) are presented Fig. 5A.

YAP expression was high in the nucleus of basal cells and the cytoplasm of luminal cells, but ARHGAP29 expression was high in the cytoplasm of both cells. Notably, YAP expression was unrelated to the Gleason score. The characteristics of the prostate cancer patients are presented in Table I. ARHGAP29 expression was significantly associated with the risk classification of prostate cancer (Fig. 5B). Both YAP and ARHGAP29 had low area under the curve (AUC) scores as prognostic markers, but there was a significant difference between the expression of these proteins and biochemical progression-free survival (b-PFS: P=0.0422, and P=0.0123, respectively) (Fig. 5C and D). In addition, high expression of both proteins was significantly associated with poor prognoses (Fig. 5D). In TCGA database, YAP did not exhibit a tendency for a poor prognosis in patients with high expression. In contrast, ARHGAP29 exhibited a tendency for a poor prognosis in patients with high expression in TCGA (Fig. S1A and B). Moreover, the prognostic significance of clinicopathological parameters, including prostate specific antigen (PSA), the D'Amico risk classification, Gleason score, and pathological T category, and the expression levels of YAP and ARHGAP29 were evaluated in prostate cancer patients (Table II). As a result, high ARHGAP29 expression was a significant independent risk factor related to b-PFS in multivariate analysis (HR=2.27; P<0.05; data not shown).