A total of 70 PPC and 56 CRPC serum samples were collected from patients prior to surgery between December, 2016 and August, 2018, and 45 PPC and 36 CRPC tissue samples were collected from patients between January, 2010 and August, 2018 at The First Affiliated Hospital of Chongqing Medical University, Chongqing, China. All the participants were informed of the aims and procedures of the study. Patients with CRPC were selected according to the European Association of Urology (EAU) guidelines (23). The present study retrospectively collected clinical data, such as age, prostate-specific antigen (PSA) levels and metastatic status from medical records. The study was approved by the Ethics Committee of Chongqing Medical University and was conducted according to the principles of the Declaration of Helsinki.

All tissue samples were fixed in 10% paraformaldehyde for 12 h at 25°C. They were then embedded in paraffin and cut into 5-µm-thick sections. Immunohistochemical staining was detected using a standard immunoperoxidase staining procedure. The sections were incubated overnight at 4°C with anti-Cav-1, (1:200, cat. no. 3267; Cell Signaling Technology, Inc., Danvers, MA, USA) or anti-H-Ras (1:50, cat. no. sc-53958), anti-K-Ras (1:50, cat. no. sc-521), anti-PLCε (1:50, cat. no. sc-28402) (all from Santa Cruz Biotechnology, Inc., Dallas, TX, USA), followed by incubation with a secondary antibody (1:100, SP9000 or PV9003; ZSGB-BIO, Beijing, China) at 37°C for 1 h, then sequentially incubated with avidinbiotin complex solution (room temperature, 1 h). Protein expression was detected by coloration with diaminobenzidine (DAB) (ZLI-9018; ZSGB-BIO) buffer for 5 min, and the sections were counterstained with hematoxylin for 5 min at room temperature. All images were captured using a fluorescent microscope (Nikon Corp., Tokyo, Japan). The staining intensity was scored as follows: 0 (no staining), 1 (light staining), 2 (moderate staining) and 3 (strong staining). The immunoreactivity ratio was scored as follows: 0 (0% immunoreactive cells), 1 (<5% immunoreactive cells), 2 (5-50% immunoreactive cells) and 3 (>50% immunoreactive cells). The final score was defined as the sum of both parameters. Scores ≤2 denoted a negative expression, while scores ≥3 denoted a positive expression.

Serum samples were centrifuged at 1,000 × g for 10 min (room temperature) and stored at -80°C until analysis. The concentrations of Cav-1 were determined using a human Cav-1 ELISA kit (detection range, 0.15-10 ng/ml; Signalway Antibody LLC, College Park, MD, USA), according to the manufacturer’s instructions. The absorbance was measured at a wavelength of 450±2 nm to detect the concentration of Cav-1 in serum using a microplate reader (Bio-Rad Laboratories, Inc., Hercules, CA, USA).

The human prostate cancer cell lines, LNCaP (CRL-1740™), PC3 (CRL-1435™) and DU145 (HTB-81™), were obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA). Castration-resistant derivatives of LNCaP cells [specifically, bicalutamide-resistant cells (Bic-R) and enzalutamide-resistant cells (En-R)] were constructed and maintained as previously described (24). All cells were cultured in RPMI-1640 medium (Thermo Fisher Scientific, Inc., Waltham, MA, USA) containing 1% penicillin and streptomycin antibiotics (Beyotime Institute of Biotechnology, Haimen, China) and 10% fetal bovine serum (Thermo Fisher Scientific, Inc.) at 37°C in a humidified atmosphere of 5% CO₂.

For transfection, 1×10⁵ cells were seeded in a 6-well plate and passaged every 2 days. When appropriate, i.e., at 1 day after seeding or when the cell cultures were 40-60% confluent, the cells were transfected with 4 µg Cav-1 knockdown plasmids or negative control (Shanghai GenePharma Co., Ltd., Shanghai, China) and 10 µl Lipofectamine 2000 (Thermo Fisher Scientific, Inc.) in serum-free medium, according to the manufacturer’s instructions. The transfection medium was replaced with fresh culture medium after 6-8 h at 37°C. The transfection process of the H-RasG12V or K-RasG12V plasmids (Shanghai GenePharma Co., Ltd.) was similar to that of the Cav-1 knockdown plasmids. However, for the transduction of PLCε knockdown lentiviral particles, 1×10⁵ cells were seeded in 6-well plates and incubated with 3 ml complete medium until 40-60% confluency, and subsequently 3 µl of PLCε knockdown lentivirus or negative control lentivirus (Shanghai GenePharma Co., Ltd.) and 3 ml of Polybrene (Thermo Fisher Scientific) were added to the culture medium for 8 h. The original medium was then replaced with 6 ml fresh medium containing 1 µg/ml puromycin. The cells were harvested after 72 h and used for protein extraction, or were cultured for 48 h and used for RNA extraction or in other experiments.

RNA was extracted from all cell lines using TRIzol reagent, and 1 µg RNA was reverse transcribed into complementary DNA using the Prime Script™ RT reagent kit, according to the manufacturer’s instructions (Takara Biotechnology Co., Ltd., Dalian, China). Messenger RNA (mRNA) levels were analyzed using the SYBR PremixEx Taq™ II kit (Takara Biotechnology Co., Ltd.) and the CFX96™ Real-Time PCR Detection System (Bio-Rad Laboratories, Inc., Hercules, CA, USA). The thermocycling conditions were as previously described (24), with the exception of the annealing temperature. For Cav-1 analysis, the annealing temperature was 52°C, while for H-RasG12V and K-RasG12V analysis, the annealing temperature were 56°C and 60°C. The mRNA expression levels were calculated using the comparative 2^−ΔΔCq method (25) and GAPDH was used as a calibrator. The primers used for the human Cav-1, H-RasG12V, K-RasG12V and GAPDH genes were as follows: Cav-1 (107 bp), forward (F), 5′-CATCCCGATGGCACTCATCTG-3′ and reverse (R), 5′-TGCACTGAATCTCAATCAGGAAG-3′; H-RasG12V (129 bp), F, 5′-CTGAGGAGCGATGACGGAA-3′ and R, 5′-AGGCTCACCTCTATAGTGGG-3′; K-RasG12V (165 bp), F, 5′-AAGGCCTGCTGAAAATGACTG-3′ and R, 5′-GGTCCTGCACCAGTAATATGCA-3′; and GAPDH (258 bp), F, 5′-AGAAGGCTGGGGCTCATTTG-3′, R, 5′-AGGGGCCAT CCACAGTCTTC-3′.

Simvastatin (S1792; Selleck Chemicals, Houston, TX, USA) was activated using absolute ethanol and adjusted to a final concentration of 10 mM with PBS (pH 7.2). Methyl-β-cyclodextrin (M-β-CD; C4555; Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) was solubilized to 30 mM in PBS. Cholesterol (C3045; Sigma-Aldrich; Merck KGaA) was solubilized to 10 mM in ethanol. All these procedures were performed according to the manufacturer’s instructions and the reagents were stored at −20°C.

Total protein extraction and subcellular fractionations were performed as previously described (26,27), although certain modifications were introduced to extract the detergent-resistant fraction (DRF) from lipid-rich membranes and the detergent-soluble fraction (DSF). Cell pellets were lysed in 1% Triton X-100, 25 mM Tris (pH 8.0), 150 mM NaCl, 10 mM Na-pyrophosphate, 10 mM NaF, 3 mM EDTA, 1 mM Phenylmethanesulfonyl fluoride (PMSF), 10 µg/ml aprotinin, 1 mM benzamidine and 1 mM sodium orthovanadate for 10 min at 4°C. An ultrasonic homogenate step was included to more finely disrupt cellular membranes, and the homogenate was incubated for 30 min on ice and then centrifuged at 100,000 × g for 30 min at 4°C. The insoluble pellets (containing the DRF) and the supernatant (containing the DSF) were resuspended by boiling in SDS-PAGE sample buffer.

Western blot analysis was performed as previously described (24,28). The intensity analyses were quantified using Image-Pro Plus 6.0 software. The membranes were incubated with primary antibodies overnight at 4°C, then sequentially incubated with secondary antibodies for 2 h at room temperature. The primary and secondary antibodies used were as follows: Anti-Cav-1 (1:1,000, cat. no. 3267), anti-matrix metalloproteinase 2 (MMP2) (1:1,000, cat. no. 40994), anti-matrix metalloproteinase 9 (MMP9) (1:1,000, cat. no. 13667), anti-Snail (1:1,000, cat. no. 3879) and anti-GAPDH (1:1,000, cat. no. 5174) antibodies were obtained from Cell Signaling Technology. Anti-H-Ras (1:500, cat. no. sc-53958), anti-K-Ras (1:500, cat. no. sc-521), anti-3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase (HMGR) (1:500, cat. no. sc-271595) and anti-PLCε (1:500, cat. no. sc-28402) antibodies were obtained from Santa Cruz Biotechnology. Goat anti-mouse IgG (1:3,000, cat. no. SA00001-1), goat anti-rabbit IgG (1:3,000, cat. no. SA00001-2) and rabbit anti-goat IgG (1:3,000, cat. no. SA00001-4) were obtained from ProteinTech (Rosemont, IL, USA).

The cells (2,000 cells/well) examined for cell viability using the CCK-8 assay were seeded in a 96-well plate and incubated for 12 h at 37°C. Following 24-72 h of treatment with various agents or concentrations, CCK-8 reagent solution (10 µl; Beyotime Institute of Biotechnology) was added to each well followed by incubation for 2 h at 37°C. The absorbance was measured at 450 nm using a microplate reader (Bio-Rad Laboratories, Inc.). Cell viability data were calculated as percentage values for each treatment condition compared with the control. The synergistic effects of simvastatin and AR antagonists was calculated using CalcuSyn^® software version 2.0 (Biosoft, Cambridge, UK), and the combination index (CI) quantitatively depicted additive effects (CI=1), synergism (CI<1) or antagonism (CI>1).

For the invasion assay, the cells (1×10⁴ cells/well) incubated in serum-free medium were added to the upper chamber of the insert with Matrigel (BD Biosciences, San Jose, CA, USA). Following incubation at 37°C for 24 h, permeable cells were fixed with 4% paraformaldehyde for 20 min and stained with 0.1% crystal violet for 10 min at room temperature. The cells were counted under an inverted microscope (Nikon Corp.) in 5 different visual fields and averaged.

For the wound-healing assay, the cells (5×10⁴ cells/well) were seeded into 6-well plates and incubated in serum-free medium for 24 h at 37°C. The cell monolayer was scratched with a 200-µl pipette tip to form wound gaps. The cells were then washed in PBS and then incubated for 24 h at 37°C continuously, and images were captured under an inverted microscope (Nikon Corp.) at the indicated time-points.

The cells (1×10⁵ cells/well) plated into 6-well plates covered with glass were incubated overnight at 37°C. Once the cells adhered to the walls, they were fixed with 4% paraformaldehyde for 20 min, the experiment was then performed as previously described (24). The cells were then incubated with the primary antibody (anti-Cav-1; 1:200; cat. no. 3267; Cell Signaling Technology, Inc.) overnight at 4°C, and then with a secondary antibody (FITC conjugated goat anti-rabbit IgG, 1:100, ZF0311; ZSGB-BIO) in a dark room at 37°C for 30 min. Cell nuclei were stained with DAPI (ZLI-9557; ZSGB-BIO) at 37°C for 5 min. Immunofluorescence images were obtained upon incubation in 50% glycerol using a fluorescence microscope (Nikon Corp.).

All experiments were repeated ≥3 times independently with technical replicates and analyzed using SPSS software version 21.0 (IBM Corp., Armonk, NY, USA) and GraphPad Prism software version 5 (GraphPad Software, Inc., La Jolla, CA, USA). Significant differences among experimental groups were evaluated using the Student’s t-test, one-way analysis of variance (ANOVA) and two-way ANOVA, while the Bonferroni adjustment was employed for multiple comparisons. Survival analyses were conducted using the Kaplan-Meier method. The Cox proportional hazards model was used to estimate hazard ratios (HRs). Other data were analyzed using Spearman’s correlation analysis, the Chi-square test and Mann-Whitney test. One-way ANOVA was used for multiple comparisons followed by the Student-Newman-Keuls test as a post hoc test. A value of P<0.05 was considered to indicate a statistically significant difference.

To determine the differential expression levels of Cav-1, H-Ras, K-Ras and PLCε in the tissues of patients with PPC and CRPC, and to explore the association between these proteins, 45 PPC and 36 CRPC tissue samples were collected from patients. Immunohistochemical assays revealed that the expression of Cav-1, H-Ras, K-Ras and PLCε was upregulated in the CRPC compared with the PPC tissue samples (P=0.002, P=0.0272, P=0.0377 and P=0.0364, respectively) (Fig. 1A and B). Spearman’s correlation analysis revealed that there was a positive correlation between the Cav-1 expression levels and the H-Ras (r=0.4151, P=0.0118), K-Ras (r=0.3443, P=0.0398) and PLCε expression levels (r=0.4338, P=0.0082) in CRPC (Fig. 1C). The demographic and clinical characteristics of these patients and the association of these characteristics with Cav-1 expression are summarized in Table I. The data revealed that 72.2% (26/36) of the CRPC samples had a positive Cav-1 expression vs. 31.1% (14/45) of the PPC samples (Table I and Fig. 1B). Among the various clinical parameters, only bone metastasis was positively associated with the expression of Cav-1 in the PPC (P=0.021) and CRPC (P=0.006) tissues, suggesting that Cav-1 overexpression may promote bone metastasis (Table I).

Kaplan-Meier survival analysis revealed that the median recurrence-free survival (RFS) was 36 months [95% confidence interval (CI) = 24-48 months] in patients with Cav-1-negative CRPC, while the median RFS was 25 months (95% CI = 22-28 months) in Cav-1-positive patients. Cav-1-positive tumor tissues were associated with a shorter RFS in patients with CRPC (P=0.013, log rank test) (Fig. 2). In univariate Cox proportional hazards regression analyses, Cav-1 was found to be significantly associated with recurrence following ADT (HR = 2.65, 95% CI = 1.162-6.029; P=0.02) (Table II), suggesting that Cav-1 is an independent risk factor for the occurrence of CRPC, and that the risk of CRPC occurring in Cav-1-positive patients was 2.65-fold greater than in Cav-1-negative patients.

To gain further insight into the functions of Cav-1 in CRPC, the Cav-1 levels in serum samples from 70 patients with PPC and 56 patients with CRPC were determined by ELISA. Statistical analysis revealed that the mean Cav-1 expression level in the serum of patients with CRPC was higher than in the serum of patients with PPC (CRPC, 1.57±0.83 ng/ml vs. PPC, 0.64±0.25 ng/ml; P<0.001) (Table III and Fig. 3A). Subsequently, the present study explored whether Cav-1 can be used as a possible diagnostic factor in CRPC by assessing receiver operating characteristic (ROC) curves. As shown in Fig. 3B, the area under the curve (AUC) was 0.876 (95% CI = 0.813-0.939), suggesting that Cav-1 could be used for the accurate diagnosis of CRPC. The diagnostic properties using different Cav-1 cut-off values are shown in Fig. 3C.

Since different cut-off values result in different values of sensitivity and specificity, the cut-off was set at 0.75 ng/ml, as it maximized the Youden index, resulting in sensitivity and specificity rates of 78.6 and 87.1%, respectively. However, considering that our study population receiving ADT is known to be at high risk of castration resistance, achieving a minimum of 80% sensitivity and maximizing sensitivity over specificity were deemed desirable. Using this criterion, the optimal cut-off value was set at 0.68 ng/ml, which could improve sensitivity to 82.1% while maintaining specificity at 80%. Consequently, Cav-1 levels were statistically evaluated with certain clinical parameters in patients with CRPC. The data revealed that an increased Cav-1 expression was positively associated with tumor metastasis (P=0.003; Table IV). Taken together, these data suggest that Cav-1 could be used as a potential biomarker to predict metastasis in patients with CRPC.

Although we demonstrated shown that the Cav-1 levels were higher in serum and tumor tissues from patients with CRPC than those with PPC, there is little evidence to indicate whether this increase regulates the progression or metastasis of CRPC. To address this question, the present study examined the mRNA and protein expression levels of Cav-1 by RT-qPCR and western blotting, respectively, in an androgen-sensitive LNCaP cell line and in castration-resistant Bic-R, En-R, PC3 and DU145 cell lines. The protein level of Cav-1 was markedly low in the LNCaP cells, but was considerably higher in the PC3, DU145, Bic-R and En-R cells. At the mRNA level, similar results were obtained (Fig. 4A-C). This finding suggested that Cav-1 expression was increased in the CRPC cells. Subsequently, to determine the relevance of the increased Cav-1 expression on the metastasis of castration-resistant tumors, a Cav-1 knockdown plasmid was constructed and transfected into the PC3 and En-R cells (Fig. 4D and E). The results of western blot analysis revealed that Cav-1 knockdown downregulated the expression levels of factors associated with invasion and migration, such as Snail, MMP2 and MMP9 (Fig. 4F-I). Transwell and wound-healing assays consistently demonstrated that the knockdown of Cav-1 suppressed the invasion and migration of CRPC cells (Fig. 4J-N).

As stated above, the isoforms of Ras, H-Ras and K-Ras and the downstream gene PLCε were overexpressed in CRPC and exhibited a positive correlation with Cav-1 expression. Therefore, we hypothesized that Cav-1 may regulate Ras signaling in cell membranes and may be associated with CRPC metastasis.

To explore this concept, a series of Cav-1 knockdown experiments were performed and the expression of H-Ras, K-Ras and PLCε in DRFs was examined. H-Ras, K-Ras and PLCε were detected in DRFs, which are composed of lipid rafts and caveolae, as confirmed by the presence of Cav-1. The non-classical localization of PLCε was specifically observed in caveolae and rafts by western blot analysis. Additionally, the knockdown of Cav-1 downregulated the expression of H-Ras and PLCε, but did not alter the expression of K-Ras in DRFs (Fig. 5A-D).

To further elucidate the role of Cav-1 in regulating PLCε through H-Ras, the cells were treated with H-RasG12V and K-RasG12V (Fig. 5E and F), which are the activated forms of H-Ras and K-Ras, respectively. The results revealed that H-RasG12V upregulated the expression of PLCε in DRFs in the different CRPC cells, and that this upregulation was reversed by the knockdown of Cav-1 (Fig. 5G and I). K-RasG12V also increased the expression of PLCε, although the knockdown of Cav-1 did not reverse this process (Fig. 5H and J). These results suggest that Cav-1 knockdown attenuates the expression of PLCε in the plasma membrane, particularly in lipid rafts, through H-Ras.

Our previous study suggested that PLCε induced the invasion and migration of bladder cancer cells (21); however, it remains unclear as to whether PLCε, in particular PLCε in DRFs, is responsible for the invasion and migration of prostate cancer. Therefore, in this study, PLCε was knocked down using lentiviral transduction particles. As shown in Fig. 6A-D, the knockdown of PLCε in PCa cells decreased the expression of factors involved in invasion and migration, such as Snail, MMP2 and MMP9, suggesting that PLCε regulates invasion and migration in CRPC.

To determine whether the process of metastasis is associated with the expression of PLCε in DRFs, PLCε was knocked down in combination with the knockdown of Cav-1. The results revealed that, compared with the knockdown of PLCε alone, the concurrent knockdown of Cav-1 downregulated the expression of PLCε only in DRFs and slightly upregulated the expression of PLCε in DSFs (although the difference was not statistically significant) (Fig. 6E-H). Consistently, the knockdown of Cav-1 in combination with the knockdown of PLCε downregulated more effectively the changes in downstream factors of invasion and migration compared with knockdown of PLCε alone (Fig. 6A-D). Using Transwell assays, consistent results were obtained (Fig. 6I-K), suggesting that the knockdown of Cav-1 suppressed invasion and migration in CRPC through the attenuation of the expression of PLCε in plasma membrane caveolae.

Murata et al have previously reported that Cav-1 can directly binds cholesterol, and that a threshold level of membrane cholesterol is required for caveolae to form (29). To delineate whether cholesterol is responsible for the expression of Cav-1, the cells were cultured in serum-free conditions, which removed the effects of exogenous cholesterol, and treated the cells with simvastatin, M-β-CD and cholesterol. Simvastatin is an inhibitor of HMGR, which causes the blockade of cholesterol biosynthesis at the rate-limiting step (HMG-CoA conversion to mevalonate) (30). M-β-CD is the most effective agent for stripping cholesterol from the cell membrane. Both drugs can therefore deplete cholesterol through different pathways.

In this study, simvastatin decreased the expression of Cav-1 in DRFs, although the level varied in different cells, similar to what was observed with M-β-CD. Inversely, cholesterol replenishment resulted in an increased membrane Cav-1 expression compared with the untreated cells (Fig. 7A and B). The immunofluorescence assay revealed similar results (Fig. 7C). Taken together, these results suggested that there was a functional association between free membrane cholesterol and Cav-1 expression in CPRC cells.

To further delineate the association between simvastatin and Cav-1, western blot analysis revealed that 10 µM simvastatin decreased the expression of HMGR and Cav-1 in En-R cells, with an even greater suppression when 20 µM simvastatin was used. A statistically significant decrease in the expression of HMGR and Cav-1 also occurred when the PC3 cells were cultured with 20 µM simvastatin, but not with 10 µM simvastatin (Fig. 7D and E). Taken together, these results indicate that simvastatin can markedly suppress membrane Cav-1 expression through inhibition of de novo cholesterol synthesis in CRPC cells.

To further explore whether Cav-1 reverses the sensitivity of the cells to AR antagonists, Cav-1 was knocked down in Bic-R and En-R cells. Cell proliferation was not suppressed by bicalutamide in the control cells, but was suppressed in the cells in which Cav-1 was knocked down, and was further suppressed in the cells in which Cav-1 was knocked down and treated with bicalutamide. Replenishment with H-RasG12V modestly reversed the suppression of cell proliferation caused by Cav-1 knockdown or by Cav-1 knockdown combined with bicalutamide. The same results were observed with enzalutamide treatment (Fig. 8A and B), suggesting that Cav-1 can reverse the sensitivity of cells to AR antagonists partly through H-RasG12V.

Finally, to explore whether simvastatin exerts antitumor effects on CRPC, cell viability was evaluated by a CCK-8 assay in cells treated with various concentrations of simvastatin. In contrast to enzalutamide or bicalutamide treatment, in these drug-resistant cells, increasing concentrations of simvastatin caused a gradual decrease in the viability of CPRC cells, and more prominently when enzalutamide or bicalutamide was combined with simvastatin (Fig. 8C and D). Consistently, simvastatin enhanced the suppressive effect of enzalutamide on the En-R or bicalutamide on Bic-R cells, as indicated by the combination indexes (CI)<1. [En-R, CI = 0.35 at effective concentration (EC) of 50, 0.19 at EC = 75 and 0.10 at EC = 90; and Bic-R, CI = 0.38 at EC = 50, 0.21 at EC = 75 and 0.12 at EC = 90] (Table V).