Prostate tissue samples were obtained from 12 benign prostatic hyperplasia (BPH) patients, 12 localized prostate cancer patients and 12 metastatic prostate cancer patients diagnosed by biopsy or postoperative pathological examination at the Department of Urology of Xijing Hospital (Xi’an, Shaanxi, China) between January 2016 and December 2016. Mean calculated ages of the three groups were 71±5.6 years (range, 62–83 years), 67±6.4 years (range, 53–77 years) and 66±4.4 years (range, 59–72 years), respectively. Written informed consent and permission to collect prostate tissues were obtained from all subjects enrolled in the study. All analyses of human materials were approved by the Research Ethics Committee of the Fourth Military Medical University (Xi’an, Shaanxi, China), and conducted according to the principles of the Declaration of Helsinki. Tissue samples were fixed with a concentration of 10% formalin at room temperature overnight, paraffin-embedded and cut into 4-µm thick sections for further staining analysis. Tissue samples that were used for reverse transcription-quantitative polymerase chain reaction (RT-qPCR) were stored at −80°C until RNA isolation.

The human prostate cancer PC3 cell line was obtained from the Type Culture Collection of the Chinese Academy of Science (Shanghai, China), and was cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (both Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA) in a humidified 5% CO₂ incubator at 37°C. Briefly, PC3 cells at 60–80% confluency were stimulated with 0.5, 1 or 2 µg/ml recombinant (r) HMGB1 (Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) for 24 h at 37°C. For the RAGE inhibition experiment, RAGE neutralizing antibody (R&D Systems, Inc., Minneapolis, MN, USA) was used at 20 µg/ml for 24 h at 37°C.

Cells were seeded into 6-well plates at 300,000 cells/well and incubated for 24 h in a humidified 5% CO₂ incubator at 37°C to form cell monolayers. Next, a sterile 10- or 100-µl plastic pipette tip was used to create a scratch wound across the monolayer, following which the detached cells were removed by washing with cold phosphate-buffered saline [PBS; 137 mM NaCl, 2.68 mM KCl, 8.10 mM Na₂HPO₄ and 1.47 mM KH₂PO₄ (pH 7.4)] three times. Cells were then cultured in serum-free DMEM with/without rHMGB1 stimulation in a humidified 5% CO₂ incubator at 37°C for 24 h, and the cells that had migrated into the wound area were imaged with a phase-contrast microscope (Nikon Corporation, Tokyo, Japan). The area of the wound at 24 h, as compared with the initial wound area, was calculated using the Fiji package of ImageJ (National Institutes of Health, Bethesda, MD, USA).

Cell migration and invasion assays were performed using Transwell chambers (pore size, 8 µm; Costar; Corning Incorporated, Corning, NY, USA), whereby the base was coated with 1 mg/ml BD Matrigel matrix (BD Biosciences, Franklin Lakes, NJ, USA) for the invasion assay. Cells were diluted at a cell density of 2×10⁵ cells/ml. A total of 200 µl DMEM containing rHMGB1 was added to the upper chamber and 600 µl DMEM was added to the lower chambers. After 24 h of culture, the cells that invaded into the lower chamber were fixed with 4% formalin for 15 min at room temperature. The cells on the inner layer were then gently removed with a cotton swab and the adherent cells on the lower surface of the insert was stained with 0.1% crystal violet for 30 min at room temperature. The filters were washed with water several times and images were captured (original magnification, ×100). Invaded cells on the lower surface were counted under a light microscope (Nikon Corporation).

The small interfering RNAs (siRNAs) for RAGE, NF-κB p65 and normal control siRNA (si-NC) were synthesized by Guangzhou RiboBio Co., Ltd. (Guangzhou, China). PC3 cells were transfected with siRNA (5 nM) and Lipofectamine^® 3000 (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer’s protocol. Subsequent to 48 h of culture at 37°C, control cells were left untreated, and rHMGB1 at 1 µg/ml was added to the remaining wells for stimulation.

For immunohistochemistry and immunofluorescence staining of prostate samples, 4-µm sections of paraffin-embedded tissues were blocked at room temperature with 5% goat serum in PBS for 30 min. Sections were then incubated with rabbit anti-HMGB1 (1:100; catalog no. ab18256; Abcam, Cambridge, UK) or anti-RAGE (1:50; catalog no. sc-365154; Santa Cruz Biotechnology, Inc., Dallas, TX, USA) at 4°C overnight, followed by a horseradish peroxidase-labeled goat anti-rabbit antibody (1:200; catalog no. CW0156; CWBio, Beijing, China) or a Cy3-conjugated goat anti-rabbit antibody (1:200; catalog no. ab6939; Abcam) for 1 h at room temperature. DAB developing system (Gene Tech Biotechnology Co., Ltd., Shanghai, China) was used for detecting the biotinylated antibodies at room temperature for 30 sec. Immunohistochemistry images were acquired with an Olympus VS120 slide scanner (Olympus Corporation, Tokyo, Japan). Two pathologists who were blinded to the clinical characteristics of the patients scored HMGB1 expression independently. As previously reported (20), the intensity of HMGB1 staining was expressed as four grades: 0, None; 1, weak; 2, moderate; and 3, strong. The percentage of HMGB1-positive cells was expressed as five grades: 0, <5%; 1, 6–25%; 2, 26–50%; 3, 51–75%; and 4, >75%. The score of the HMGB1 expression was evaluated by multiplying the intensity by the percentage. In addition, confocal images were acquired using the Olympus Fluoview 1000 microscope (Olympus Corporation) with a PlanApo N lens (×40 with and without ×2.5 digital zoom).

For cell immunofluorescence staining, PC3 cells were seeded and incubated at 37°C, in 5% CO₂, followed by stimulation with rHMGB1 (1 µg/ml) for 24 h. Cells were then washed with ice-cold PBS and fixed immediately with 4% paraformaldehyde for 15 min at room temperature, then permeabilized with 0.1% Triton X-100 for 15 min. Cells were incubated with anti-E-cadherin (catalog no. 14472) or vimentin (catalog no. 5741) (both 1:100; Cell Signaling Technologies, Inc., Danvers, MA, USA), or fibronectin 1 (1:100; catalog no. ab2413; Abcam) overnight at 4°C. Cy3-conjugated goat anti-rabbit antibody (1:200; catalog no. ab6939; Abcam) was used as the secondary antibody. Nuclear DNA was detected by incubating cells with Hoechst 33342 (1:1,000; Sigma-Aldrich; Merck KGaA) for 20 min at room temperature.

The total RNA from the tissues and cells was extracted using TRIzol (Takara Bio, Inc., Otsu, Japan), and cDNA was synthesized using reverse transcriptase from a Perfect Real-Time PrimeScript™ RT Reagent kit (Takara Bio, Inc.). The relative content of mRNA was amplified via qPCR with SYBR-Green (SYBR^® Premix Ex Taq™ II; Takara Bio, Inc.) on a Chromo4 continuous fluorescence detector with a PTC-200 DNA Engine Cycler (Bio-Rad Laboratories, Inc., Hercules, CA, USA). Thermal cycling conditions were as follows: Initial denaturation for 30 sec at 95°C, followed by 40 cycles of denaturation for 5 sec at 95°C and extension for 30 sec at 60°C; the absorbance value was read at the extension stage. All values were normalized to the expression of β-actin. The primers used are listed in Table I. Relative quantification was performed according to the ΔΔCq method, and results were expressed in the linear form using the formula 2^−ΔΔCq (23). Results were considered significant when at least a 2-fold difference in expression levels was detected.

Western blot assays were performed using whole cell lysates from cultured PC3 cells. Cells were washed twice with ice-cold PBS and lysed in 2X sodium dodecyl sulfate (SDS) sample buffer (2% mercaptoethanol, 20% glycerol, 4% SDS in 100 mM Tris-HCl buffer, pH 6.8) containing a protease inhibitor cocktail (Roche Diagnostics, Basel, Switzerland) and a phosphatase inhibitor cocktail (Thermo Fisher Scientific, Inc.), and incubated for 15 min on ice. The supernatants were collected by centrifugation at 12,000 × g for 15 min at 4°C, and a BCA protein assay kit was used to measure the protein content. Protein samples (20 µg) were loaded and separated with 10% SDS-PAGE, and then transferred onto polyvinylidene difluoride membranes. Subsequently, membranes were blocked with Tris-buffered saline-Tween-20 buffer containing 5% skimmed milk for 1 h at room temperature. Membranes were incubated at 4°C overnight using the following antibodies: RAGE (catalog no. sc-365154; Santa Cruz Biotechnology, Inc.), NF-κB p65 (catalog no. 8242), phospho-NF-κB p65 (catalog no. 3033), E-cadherin (catalog no. 3195), N-cadherin (catalog no. 13116), vimentin (catalog no. 5741) (all from Cell Signaling Technology, Inc.) and fibronectin 1 (catalog no. ab2413; Abcam), with β-actin (catalog no. CW0264; CWBio) as an internal control. The membranes were then developed with an enhanced chemiluminescence detection kit (Advansta, Inc., Menlo Park, CA, USA).

Cells were trypsinized and seeded at 3,000 cells per well in 96-well plates. Cell proliferation was then measured using the CCK-8 kit (catalog no. C0038; Beyotime Institute of Biotechnology, Shanghai, China) every 24 h according to the manufacturer’s protocols. Absorbance was detected at a wavelength of 450 nm. Three wells were measured for cell viability in each treatment group.

Unless stated otherwise, all experiments were performed three times. All statistical analyses were performed using GraphPad Prism version 6 software (GraphPad Software, Inc., La Jolla, CA, USA). Data were analyzed using the Student’s unpaired two-tailed t-test when two groups were compared. Statistical significance was also determined using analysis of variance where indicated in the figure legends. The number of sampled units (n) is also indicated in the figure legends. P<0.05 was considered to indicate a statistically significant difference.

To examine the expression of HMGB1 in the tissues of BPH, localized prostate cancer and metastatic prostate cancer, immunohistochemical staining and RT-qPCR were performed. As the immunochemistry results showed, the protein level of HMGB1 was markedly high in localized prostate cancer tissues and metastatic prostate cancer tissues, and less detectable in BPH tissues, with H-Scores of 228.8±12.88, 269.4±7.223 and 105.6±12.94, respectively (Fig. 1A and B). These results were confirmed by RT-qPCR, which demonstrated significantly elevated HMGB1 mRNA in localized and metastatic prostate cancer tissues compared with that in BPH tissues (Fig. 1C). Immunohistochemical staining and RT-qPCR results indicated a moderate and significant elevation, respectively, in the HMGB1 levels in metastatic prostate cancer tissues compared with that in primary prostate cancer tissues (Fig. 1B and C). Consistent with previous studies, these findings indicated that HMGB1 was overexpressed in prostate cancer tissues, and particularly predominant in metastatic prostate cancer tissues.

The study next investigated whether HMGB1 was involved in the invasion and metastasis of PC3 cells using a wound-healing assay, as well as Transwell migration and invasion assays. Following stimulation by recombinant HMGB1 (rHMGB1) at 0.5, 1 and 2 µg/ml, PC3 cells exhibited gradually increased migratory capacity in the wound-healing and Transwell assays (Fig. 2A and B). In the Transwell experiments aimed at detecting PC3 cell invasion, the stimulating effects of 1 µg/ml rHMGB1 on PC3 cells were most evident compared with that in the other groups, inducing significant PC3 cell invasion through the Matrigel matrix (Fig. 2C). The quantitative analyses of the migration and invasion of the cells are also presented in Fig. 2. Taken together, these data revealed that HMGB1 enhanced the metastatic and invasive capabilities of PC3 cells.

Next, the underlying mechanism of HMGB1-induced migration and invasion in PC3 cells was investigated. EMT is recognized as an essential process that is involved in cancer metastasis by providing tumor cells with enhanced mobility (15). Therefore, the expression levels of EMT markers were evaluated by RT-qPCR, western blot analysis and cell immunofluorescence. As the CCK-8 assay results showed, rHMGB1 at 2 µg/ml enhanced the proliferation of PC3 cells in a time-dependent manner compared with that in groups of lower concentrations (data not shown). PC3 cells treated with rHMGB1 at different concentrations showed significant changes in epithelial markers in the 1-µg/ml group (data not shown). Therefore, rHMGB1 at 1 µg/ml was chosen as the optimal stimulus to investigate the mechanism underlying HMGB1-induced migration and invasion.

In the present study, it was demonstrated that treatment of PC3 cells with 1 µg/ml rHMGB1 for 24 h led to a significant decrease in the mRNA expression of epithelial markers, including E-cadherin and VDR, and an evident increase in the expression levels of mesenchymal markers, including N-cadherin, vimentin, CTGF and fibronectin 1 (Fig. 3A). Furthermore, western blot analysis demonstrated the down-regulation of E-cadherin and upregulation of N-cadherin, fibronectin 1 and vimentin in PC3 cells stimulated with 1 µg/ml rHMGB1 for 24 h (Fig. 3B). In addition, the cell immunofluorescence results showed that the expression levels of the mesenchymal markers fibronectin 1 and vimentin were higher in rHMGB1-treated PC3 cells compared with those in untreated normal cells, while E-cadherin ectodomain immunofluorescence was reduced (Fig. 3C). As the MMP family has long been recognized as an important mediator of tumor metastasis (24), the mRNA expression levels of several primary MMPs were detected in PC3 cells to elucidate the MMPs that were involved in HMGB1-induced metastasis. As presented in Fig. 3D, the mRNA expression levels of MMP-1, -3 and -10, but not other MMPs, in the PC3 cells were significantly upregulated by rHMGB1 stimulation (Fig. 3D). Collectively, the aforementioned findings revealed that HMGB1 induced the EMT process and overexpression of MMPs in PC3 cells, thus serving a pathogenic role in the metastasis of prostate cancer.

To define the mechanisms by which HMGB1 induced EMT to promote cancer metastasis, the receptors for HMGB1 were then detected in PC3 cells. As reported previously, extracellular HMGB1 binds to several cell surface receptors, including RAGE and toll-like receptor 2 and 4 (TLR2 and 4) (25). Therefore, RT-qPCR and western blotting were performed to examine the expression levels of possible receptors in rHMGB1-treated PC3 cells. It was demonstrated that the receptor RAGE, but not TLR2 or TLR4, was significantly upregulated (Fig. 4A and B). The expression and localization of RAGE in BPH tissues and metastatic prostate cancer tissues were then evaluated by tissue immunofluorescence, which revealed strong positive staining of RAGE in prostate cancer and stromal cells in the metastatic prostate cancer tissues, and moderate staining in the BPH tissues (Fig. 4C). To further determine whether RAGE participated in the stimulatory effects of HMGB1 on PC3 cells, siRNA targeting RAGE was applied to reduce its expression level (data not shown). The results of the wound healing and Transwell assays demonstrated that the previously elevated invasive and metastatic capabilities of rHMGB1-treated PC3 cells were markedly attenuated by RAGE siRNA (Fig. 4D and E). The rHMGB1-induced expression changes in EMT markers, including E-cadherin, VDR, N-cadherin, vimentin, CTGF and fibronectin 1, were reversed by the siRNA targeting RAGE (Fig. 4F). In addition, following the application of a neutralizing antibody against RAGE in PC3 cells, similar outcomes were observed to that of silencing the receptor using siRNA (Fig. 5A and B). As determined by CCK-8 assay, the neutralizing antibody of RAGE did not influence the viability of PC3 cells at 24, 48, 72 and 96 h (data not shown). Furthermore, the increased expression levels of MMP-1, -3 and -10 induced by rHMGB1 were significantly lowered in the cells that were transfected with the siRNA targeting RAGE or pre-treated with RAGE antibody (Figs. 4F and 5C). Collectively, these observations revealed that the HMGB1/RAGE axis contributed to the EMT process and MMP secretion, thus leading to the migration and invasion of PC3 cells.

As previously reported, the activation of different signaling pathways, including mitogen-activated protein kinase, phosphoinositide-3 kinase/protein kinase B, Smad and NF-κB, contributes to the development of EMT in various cancer types (15). To elucidate which pathway was involved in HMGB1-induced EMT in prostate cancer, the activation of the aforementioned signaling pathways was determined by western blot analysis. As presented in Fig. 6A, rHMGB1 significantly triggered NF-κB p65 phosphorylation in PC3 cells, peaking at 20 min (Fig. 6A). Other signaling pathways were not signifi-cantly activated in PC3 cells by rHMGB1 (data not shown). In addition, silencing RAGE with siRNA or blocking RAGE with neutralizing antibody reduced the phosphorylation of NF-κB in rHMGB1-stimulated PC3 cells (Fig. 6B), and indicated that rHMGB1 activated RAGE and its downstream NF-κB pathway to induce EMT in prostate cancer.

Next, NF-κB was knocked down using siRNA to identify its pathogenic role (data not shown), and the results of the wound-healing and Transwell assays demonstrated that the numbers of migrated and invasive cells were markedly decreased in rHMGB1-stimulated PC3 cells due to NF-κB silencing (Fig. 6C and D). Furthermore, siRNA targeting NF-κB reversed the mRNA expression levels of these EMT markers and MMPs in PC3 cells triggered by rHMGB1, as demonstrated by the RT-qPCR results (Fig. 6E). Therefore, we hypothesize that HMGB1 activates the RAGE/NF-κB signaling pathway to induce EMT and upregulate MMP expression, thus promoting the progression of prostate cancer.