Melatonin Inhibits Migration and Invasion in LPS-Stimulated and -Unstimulated Prostate Cancer Cells Through Blocking Multiple EMT-Relative Pathways.

PURPOSE: Gram-negative bacteria are usually found in prostate cancer (PCa) tissues. This study aims to investigate the role of lipopolysaccharide (LPS), a glycolipid compound found in the outer membrane of gram-negative bacteria, on the migration and invasion of PCa cells, and to evaluate the protective effect of melatonin.
MATERIALS AND METHODS: DU145, PC-3 and LNCaP cells were incubated with LPS in the presence or absence of melatonin. Wound healing and Transwell assays were used to analyze migration and invasion of PCa cells. RT-PCR and Western blotting were used to assess the mRNA and protein levels, respectively. Co-IP was used to analyze β-catenin ubiquitination.
RESULTS: Our results showed that LPS promoted migration and invasion of PCa cells. In addition, LPS stimulated inflammatory reaction and induced epithelial-mesenchymal transition (EMT) in PCa cells by activating several TLR4 downstream pathways. Specifically, LPS promoted NF-κB/IL-6/STAT3 signal transduction. In addition, LPS upregulated phosphorylation levels of cytoplasmic AKTSer473 and GSK-3βSer9. Moreover, LPS induced phosphorylation of GSK-3βSer9 in the "disruption complex", and then inhibited phosphorylation and ubiquitination of cytoplasmic β-catenin, leading to β-catenin nuclear translocation. Interestingly, melatonin inhibited invasion and migration not only in LPS-stimulated but also in LPS-unstimulated PCa cells. Melatonin suppressed PCa cells migration and invasion by blocking EMT mediated by IL-6/STAT3, AKT/GSK-3β and β-catenin pathways.
CONCLUSION: This study provides evidence that melatonin inhibits migration and invasion through blocking multiple TLR4 downstream EMT-associated pathways both in LPS-stimulated and -unstimulated PCa cells. Our results provide new insights into the role of bacterial infection in PCa metastasis and a potential therapeutic agent.

Introduction

Prostate cancer (PCa) is one of the leading causes of cancer mortality in men, especially in Western developed countries.1 The high mortality of PCa is mainly due to metastasis during cancer progression.2 Chronic inflammation is often observed in prostate cancer tissues and is considered to be a driver of PCa progression, and one of the important etiology for intra-prostatic inflammation attributes to bacterial infection.1–5 In fact, Gram-negative bacteria are usually found in PCa tissue.6 Lipopolysaccharide (LPS) is the main component of the outer membrane of gram-negative bacteria, which activates immune cells to produce a variety of cytokines and subsequent immune and inflammatory responses by binding to Toll-like receptor 4 (TLR4).7 Increasing evidence indicates that TLR4 expression is elevated in PCa tissues, and activation of LPS/TLR4 signaling can promote prostate cancer progression.7 LPS has been reported to enhance the metastasis in several types of tumors, including PCa.2,8–10 However, the pathologic role of LPS in PCa metastasis remains elusive. It is well known that epithelial–mesenchymal transition (EMT) plays a key role in tumor metastasis.11,12 Several key EMT-associated pathways, including IL-6/STAT3, AKT/GSK-3β and Wnt/β-catenin, are over-activated in PCa.13–15 We hypothesized that activation of these pathways is associated with increased migration and invasion capability in LPS-stimulated PCa cells.

Melatonin, a hormone synthesized primarily by the pineal gland, is well known for regulating circadian rhythms. In addition, melatonin exerts a variety of physiological functions, such as anti-oxidant, anti-inflammatory, anti-tumor and immune-modulation, by G-protein coupled MT1 and MT2 receptors-dependent and -independent ways.16–19 As a promising anticancer agent, increasing data demonstrate that the pharmacological dose of melatonin exerts multidimensional oncostatic effects throughout all stages of tumor initiation and progression.20 Nevertheless, whether melatonin curbs the migration and invasion in LPS stimulated PCa cells remains to be determined.

The objectives of the present study were to investigate the effects of LPS on PCa cells migration and invasion, and evaluate the protective effects of melatonin. Our results provided evidence that LPS promoted migration and invasion of PCa cells. In addition, LPS induced EMT of PCa cells through activating TLR4 downstream IL-6/STAT3, AKT/GSK-3β and β-catenin pathways. Melatonin inhibited migration and invasion both in LPS-stimulated and -unstimulated PCa cells. Overall, we provided evidence that melatonin inhibits migration and invasion by blocking multiple TLR4 downstream EMT-associated pathways not only in LPS-stimulated but also in LPS-unstimulated PCa cells.

Materials and Methods

Chemicals and Reagents

Melatonin and Lipopolysaccharide (Escherichia coli LPS, serotype 0127:B8) were purchased from Sigma Chemical Co. (St Louis, MO). Antibodies against E-cadherin (#14472), Vimentin (#5741), N-cadherin (#13116), Snail (#3895), Slug (#9585), Twist (#46702), β-catenin (#8480), p-β-catenin (#9561), AKT (#4691), p-AKTSer473 (#4060), GSK-3β (#5676), p-GSK3βSer9 (#9323), p-STAT3 (#9145) β-actin (#4970), Ubiquitin (#3933) and Lamin A/C (#2032) were purchased from Cell Signaling Technology (Beverley, MA). Zeb1 (#ab203829), Axin (#ab233652), and I-κB (#ab32518) antibodies were purchased from Abcam (Cambridge, MA). Antibodies against p-IκB (#sc8404) and STAT3 (#sc8019) were from Santa Cruz Biotechnologies (Santa Cruz, CA). ECL detection kit was from Pierce Biotechnology (Rockford, IL). All the other reagents were purchased from Sigma or as indicated in the specified method.

Cell Culture and Treatments

One androgen-sensitive human PCa cell line (LNCaP) and two androgen-insensitive cell lines (PC-3 and DU145) were obtained from the Institute of Cell Research, Chinese Academy of Sciences (Shanghai, China). PC-3 cells were cultured in F-12 Ham’s medium (HyClone) containing streptomycin (100 μg/mL), penicillin (100 U/mL) and 10% fetal bovine serum (FBS, Gibco) under 5% CO2, 37°C environment. DU145 cells were cultured in MEM medium (Invitrogen) containing streptomycin (100 μg/mL), penicillin (100 U/mL), sodium pyruvate (1 mM, Gibco), GlutaMAX (1%, Gibco), and 5% FBS under 5% CO2, 37°C environment. LNCaP cells were cultured in RPMI 1640 medium (HyClone) containing streptomycin (100 μg/mL), penicillin (100 U/mL) and 10% FBS under 5% CO2, 37°C environment. About 70% confluent, the medium was replaced with serum-free medium. After a 6 h incubation, the cells were incubated with LPS (2 μg/mL) for different times in the absence or presence of melatonin (0.5, 1, or 2 mM). The cells were harvested for Western blotting, real-time PCR, and Co-immunoprecipitation (Co-IP) assays at indicated time.

Wound Healing Assay

Wound healing assays were performed in accordance with the previous method, with slight modifications.21 In brief, PC-3 cells (2.0×105 cells/well) were cultured in six-well plate until 80% confluent. The monolayer cells were replaced with fresh serum-free medium for 12 h and then carefully scratched using a 200 μL tip and washed twice with PBS. After pre-treatment with melatonin (2 mM) for 0.5 hours, then add LPS (2 μg/mL) to the corresponding wells. Cells were imaged at low magnifications for time intervals of 0, 12 and 24 hours under a light microscope. The wounded area was calculated according to the following formula: (mean wounded breadth − mean remained breadth)/mean wounded breadth×100 (%).

Cell Migration and Invasion Assays

As mentioned earlier, the migration and invasion ability of DU145 and PC-3 cells was evaluated using 8-μm Transwell filter membrane (Costar Corning).21 In the migration assay, 4×104 cells were seeded in the upper compartment supplemented with 200 μL complete medium. After the cells were cultured for 12 hours adhering to the polycarbonate membrane, replace the medium in the upper and lower chambers with serum-free medium (containing 2 mM melatonin and/or 2 μg/mL LPS). After 18 h incubation, add 60 μL FBS (10%) to the lower chamber as a chemoattractant and then incubate for 6 hours. In the invasion assay, 1.0×105 cells were seeded in the upper compartment supplemented with 200 μL complete medium, precoated with 50 μL Matrigel solution (BD Biosciences, San Jose, CA). After incubation for 12 h, the upper compartment medium was replaced with serum-free medium (containing 2 mM melatonin and/or 2 μg/mL LPS). The lower compartments were added to 60 μL FBS as a chemoattractant and incubated for 48 h. Wipe the non-migrating cells on the upper side of the membrane carefully with a wet cotton swab. The membranes were fixed with methanol for 20 minutes and stained with 0.1% crystal violet solution for 20 minutes. The number of migrating cells was evaluated by detecting the absorbance of decolorization solution of 30% acetic acid at 570 nm.

Isolation of Total RNA and Real-Time RT-PCR

Total RNA was extracted from PC-3 cells using TRI reagent according to the manufacturer’s protocol. Then, the total RNA (2.0 μg) was reverse transcribed into cDNA using AMV (Promega). Real-time PCR was performed as mentioned before.21 The primers used in the RT-PCR experiments were synthesized by Life Technologies and listed in Table 1. The amplification reaction was performed on a LightCycler® 480 instrument (Roche Diagnostics GmbH) for 50 cycles, in a three-step process of denaturation, annealing and extension. Use Light Cycler 480 software to calculate the relative proportion of target genes.

Table 1

Oligonucleotide Sequence of Primers for Real-Time RT-PCR

Genes	Forward (5ʹ-3ʹ)	Reverse (5ʹ-3ʹ)
18S	CGGCTACCACATCCAAGGAA	GCTGGAATTACCGCGGCT
IL-6	AGACAGCCACTCACCTCTTCAG	TTCTGCCAGTGCCTCTTTGCTG
IL-8	ACCACCGGAAGGAACCATCT	AGCACTCCTTGGCAAAACTG
IL-1α	GAACGTCGAAAAGAAAAGTCTCG	CCTTATCAAGATGCGAACTCACA
IL-1β	AGCTACGAATCTCCGACCAC	CGTTATCCCATGTGTCGAAGAA
TFN-α	GAGGCCAAGCCCTGGTATG	CGGGCCGATTGATCTCAGC
COX-2	TAAGTGCGATTGTACCCGGAC	TTTGTAGCCATAGTCAGCATTGT
MCP-1	CAGCCAGATGCAATCAATGCC	TGGAATCCTGAACCCACTTCT
MMP-3	CGGTTCCGCCTGTCTCAAG	CGCCAAAAGTGCCTGTCTT
MMP-9	TGTACCGCTATGGTTACACTCG	GGCAGGGACAGTTGCTTCT
uPA	GGGAATGGTCACTTTTACCGAG	GGGCATGGTACGTTTGCTG

Western Blotting

Total, cytoplasmic and nuclear proteins were extracted as described before.21 SDS-PAGE was used to transfer the same amount of protein (10~20 μg) to the PVDF membrane. The membranes were incubated for 2 h with following antibodies: E-cadherin (1:2000), N-cadherin (1:1000), Vimentin (1:2000), Zeb1 (1:1000), Snail (1:1000), Slug (1:1000), Twist (1:1000), β-catenin (1:2000), p-β-catenin (1:1000), p-IκB (1:2000), I-κB (1:2000), p-AKTSer473 (1:1000), AKT (1:2000), p-GSK3βSer9 (1:2000), GSK3β (1:2000), p-STAT3 (1:1000), STAT3 (1:2000) and Axin (1:1000). After incubated with goat anti-rabbit IgG or goat anti-mouse antibody for 1~2 h, use the ECL detection kit to develop the signal of the PVDF membrane.

Co-Immunoprecipitation (Co-IP) Assay

The immunoprecipitation assay was performed as described before.21 In brief, preparation of cytoplasmic protein of DU145 cells using lysis buffer (0.1% NP-40, 150 mM NaCl, and 50 mM Tris-HCl, pH 7.5). Cytoplasmic proteins (800 μg) were pretreatment with protein A/G-agarose (Santa Cruz) and then incubated with anti-Axin or anti-β-catenin antibody at 4°C overnight. Before collecting the immune complexes captured by agarose, wash the pellet with cold non-denaturing lysis buffer. The immune complex was separated by boiling and used for subsequent Western blotting analysis.

Enzyme-Linked Immunosorbent (ELISA)

According to the manufacturer’s protocol, a commercial ELISA kit (China Beijing 4A Biotechnology Co., Ltd.) was used to determine the level of IL-6 in the culture supernatant.

Statistical Analysis

All data are expressed as mean±SEM. Use SPSS 13.0 statistical software for statistical analysis. All statistical tests used 0.05 alpha value for two-way test. Analysis of variance and Student-Newman–Keuls post-test are used to determine the differences between different groups.

Results

Melatonin Inhibits Migration and Invasion in LPS-Stimulated and -Unstimulated PCa Cells

Effects of melatonin on migration in LPS-stimulated and -unstimulated PCa cells were analyzed. Wound healing and Transwell migration assays showed that cell migration capability was enhanced in LPS-treated DU145 and PC-3 cells (Figure 1A and B). As expected, melatonin inhibited LPS-induced migration. Of interest, melatonin alleviated migration in LPS-unstimulated DU145 and PC-3 cells (Figure 1A and B). Effects of melatonin on invasion were then analyzed in LPS-stimulated and -unstimulated PCa cells. As shown in Figure 1C, the invasion capability was enhanced in LPS-treated DU145 and PC-3 cells. Moreover, MMP-3, MMP-9 and uPA mRNAs were upregulated in LPS-treated DU145 cells (). Melatonin inhibited invasion of DU145 and PC-3 cells induced by LPS (Figure 1C). Moreover, melatonin inhibited LPS-induced upregulation of MMP-3, MMP-9 and uPA mRNAs in DU145 cells (). Interestingly, melatonin suppressed invasion in LPS-unstimulated DU145 and PC-3 cells (Figure 1C). In addition, melatonin repressed mRNA levels of MMP-3, MMP-9 and uPA in LPS-unstimulated DU145 cells ().

Figure 1

Melatonin inhibits migration and invasion in LPS-stimulated and -unstimulated DU145 and PC-3 cells. After pretreated with 2 mM melatonin for 30 min, DU145 and PC-3 cells were incubated with LPS (2.0 μg/mL). (A) The migration of PC-3 cells was measured using wound healing assay. After 12 and 24 h migration, the scratches were photographed (left panel) and wounded areas were calculated (right panel). (B) The migration of DU145 and PC-3 cells were measured using Transwell migration assay. (C) The invasion of DU145 and PC-3 cells were measured using Transwell invasion assay. All experiments were repeated three times. Data were expressed as means ± S.E.M., **P<0.01 versus control group, ##P<0.01 versus LPS group.

Melatonin Inhibits EMT in LPS-Stimulated and -Unstimulated DU145 and LNCaP Cells

Effects of melatonin on EMT in LPS-stimulated and -unstimulated PCa cells were analyzed. As shown in Figures 2A and , E-cadherin was downregulated in LPS-stimulated DU145 and LNCaP cells. By contrast, vimentin and N-cadherin were upregulated in LPS-stimulated DU145 and LNCaP cells. Melatonin alleviated LPS-mediated downregulation of E-cadherin in DU145 (Figure 2B) and LNCaP cells (). In addition, melatonin suppressed LPS-induced upregulation of vimentin and N-cadherin (Figures 2B and ). Interestedly, melatonin upregulated E-cadherin while downregulated vimentin and N-cadherin in LPS-unstimulated DU145 cells in a dose-dependent manner (Figure 2C). Next, we analyzed the effects of LPS on EMT-related transcription factors in DU145 and LNCaP cells. As shown in Figures 2D and , nuclear Zeb1, Snail, Slug and Twist were elevated in LPS-stimulated DU145 and LNCaP cells. Melatonin inhibited LPS-mediated elevation of nuclear Zeb1, Snail, Slug and Twist (Figures 2E and ). Of interest, melatonin decreased nuclear Zeb1, Snail, Slug and Twist in LPS-unstimulated DU145 cells in a dose-dependent manner (Figure 2F).

Figure 2

Melatonin inhibits EMT in LPS-stimulated and -unstimulated DU145 cells. (A) DU145 cells were incubated with LPS (2 μg/mL) for indicated time. N-cadherin, E-cadherin and Vimentin were measured using Western blotting. (B) DU145 cells were incubated with LPS (2 μg/mL) with/without melatonin (2 mM) for 24 h. E-cadherin, N-cadherin, and Vimentin were measured. (C) DU145 cells were incubated with melatonin at an indicated dose for 24 h. E-cadherin, N-cadherin, and Vimentin were measured. (D) DU145 cells were incubated with LPS (2 μg/mL) for indicated time. Nuclear Zeb1, Snail, Slug and Twist were measured using Western blotting. (E) DU145 cells were incubated with LPS (2 μg/mL) with/without melatonin (2 mM) for 12 h. Nuclear Zeb1, Snail, Slug and Twist were measured. (F) DU145 cells were incubated with melatonin at indicated dose for 12 h. Nuclear Zeb1, Snail, Slug and Twist were measured. All experiments were repeated three times. Data were expressed as means ± S.E.M., *P<0.05, **P<0.01 versus control group, #P<0.05, ##P<0.01 versus LPS group.

Melatonin Inhibits the Activation of NF-κB-Mediated Inflammatory Signals in LPS-Stimulated and Unstimulated PCa Cells

Effects of melatonin on NF-κB signal transduction in LPS-stimulated and -unstimulated PCa cells were analyzed. As shown in Figure 3A, I-κB was rapidly phosphorylated in LPS-stimulated DU145 cells as early as 0.5 h. Melatonin suppressed LPS-mediated I-κB phosphorylation (Figure 3B). Interestingly, melatonin inhibited I-κB phosphorylation in LPS-unstimulated DU145 cells in a dose-dependent manner (Figure 3C). As shown in Figure 3D–I, mRNA expressions of representative inflammatory factors (IL-1α, IL-1β, TNF-α and COX-2) and chemokines (IL-8 and MCP-1) were significantly up-regulated in LPS-stimulated PC-3 cells. Melatonin attenuated LPS-induced upregulation of IL-1α, IL-1β, TNF-α, COX-2, IL-8 and MCP-1 mRNAs (Figure 3D–I). In addition, melatonin inhibited IL-1α, IL-1β, TNF-α, COX-2, IL-8 and MCP-1 mRNAs in LPS-unstimulated PC-3 cells (Figure 3D–I).

Figure 3

Melatonin inhibits activation of NF-κB mediated inflammatory signal in LPS-stimulated and -unstimulated DU145 and PC-3 cells. (A) DU145 cells were incubated with LPS (2 μg/mL) for indicated time. Phosphorylated I-κB was measured using Western blotting. (B) DU145 cells were incubated with LPS (2 μg/mL) in absence or presence of melatonin (2 mM) for 2 h. Phosphorylated I-κB was measured. (C) DU145 cells were incubated with melatonin for indicated dose. Phosphorylated I-κB was measured. (D–I) PC-3 cells were incubated with LPS (2 μg/mL) in absence or presence of melatonin (2 mM) for 6 h. (D) IL-1α; (E) IL-1β; (F) TNF-α; (G) COX-2; (H) IL-8; (I) MCP-1 mRNAs were detected using RT-PCR. All experiments were repeated three times. Data were expressed as means ± S.E.M., *P<0.05, **P<0.01 versus control group, ##P<0.01 versus LPS group.

Melatonin Inhibits STAT3 Signal Transduction in LPS-Stimulated and -Unstimulated DU145 and PC-3 Cells

Effects of melatonin on STAT3 signal transduction in LPS-stimulated and -unstimulated PCa cells were analyzed. As shown in Figure 4A, IL-6 mRNA was significantly upregulated in LPS-stimulated PC-3 cells. Correspondingly, IL-6 level in culture medium was significantly elevated in LPS-stimulated PC-3 cells (Figure 4B). Melatonin attenuated LPS-induced upregulation of IL-6 (Figure 4A and B). In addition, melatonin inhibited IL-6 expression and secretion in LPS-unstimulated PC-3 cells (Figure 4A and B). As shown in Figure 4C and D, STAT3 phosphorylation was elevated in LPS-treated DU145 and PC-3 cells (Figure 4C and D). As expected, melatonin attenuated LPS-mediated STAT3 phosphorylation. Moreover, melatonin inhibited STAT3 phosphorylation in LPS-unstimulated DU145 cells in a dose-dependent manner (Figure 4E).

Figure 4

Melatonin inhibits STAT3 signal transduction in LPS-stimulated and -unstimulated DU145 and PC-3 cells. (A and B) PC-3 cells were incubated with LPS (2 μg/mL) with/without melatonin (2 mM) for 6 h. (A) IL-6, mRNAs were detected using RT-PCR. (B) IL-6 levels in the supernatant were detected by ELISA. (C and D) PC-3 and DU145 cells were incubated with LPS (2 μg/mL) in absence or presence of melatonin (2 mM) for 6 h. STAT3 phosphorylation was measured. (E) DU145 cells were incubated with melatonin at indicated dose for 6 h. STAT3 phosphorylation was measured. All experiments were repeated three times. Data were expressed as means ± S.E.M., *P<0.05, **P<0.01 versus control group, ##P<0.01 versus LPS group.

Effects of Melatonin on AKT/GSK-3β Pathway in LPS-Stimulated and -Unstimulated DU145 Cells

We analyzed the effects of melatonin on AKT phosphorylation in LPS-stimulated and -unstimulated DU145 cells. As shown in Figure 5A, phosphorylated AKT on serine 473 (p-AKTSer473) was rapidly elevated in LPS-stimulated DU145 cells. Melatonin inhibited LPS-mediated AKTSer473 phosphorylation (Figure 5B). Of interest, melatonin downregulated p-AKTSer473 in LPS-unstimulated DU145 cells (Figure 5C). Next, the effect of melatonin on Cytoplasmic GSK-3β phosphorylation on serine 9 (p-GSK3βSer9) was analyzed. As expected, p-GSK3βSer9 was elevated in LPS-stimulated DU145 cells (Figure 5D). Melatonin inhibited LPS-mediated upregulation of p-GSK3βSer9 (Figure 5E). Interestingly, melatonin downregulated p-GSK3βSer9 in LPS-unstimulated DU145 cells (Figure 5F)

Figure 5

Effect of melatonin on AKT/GSK-3β pathway in LPS-stimulated and -unstimulated DU145 cells. (A) DU145 cells were incubated with LPS (2 μg/mL) for indicated time. Phosphorylation of AKT on serine 473 (p-AKTSer473) in cytoplasm was measured using Western blotting. (B) DU145 cells were incubated with LPS (2 μg/mL) with/without melatonin (2 mM) for 2 h. Cytoplasmic p-AKTSer473 was measured. (C) DU145 cells were incubated with melatonin for indicated dose. Cytoplasmic p-AKTSer473 was measured. (D) DU145 cells were incubated with LPS (2 μg/mL) for indicated time. Cytoplasmic p-GSK3βSer9 was measured. (E) DU145 cells were incubated with LPS (2 μg/mL) with/without melatonin (2 mM) for 2 h. Cytoplasmic p-GSK3βSer9 was measured. (F) DU145 cells were incubated with melatonin for indicated dose. Cytoplasmic p-GSK3βSer9 was measured. All experiments were repeated three times. Data were expressed as means ± S.E.M., **P<0.01 versus control group, ##P<0.01 versus LPS group.

Melatonin Inhibits β-Catenin Nuclear Translocation in LPS-Stimulated and -Unstimulated DU145 Cells

Effects of melatonin on phosphorylation and nuclear translocation of cytoplasmic β-catenin were analyzed in DU145 cells. As shown in Figure 6A, β-catenin phosphorylation was downregulated in DU145 cells as early as 0.5 h after LPS. By contrast, nuclear β-catenin was obviously increased in LPS-stimulated DU145 cells, beginning as early as 6 h after LPS (Figure 6B). Interestingly, melatonin reversed LPS-mediated downregulation of β-catenin phosphorylation (Figure 6C). In addition, melatonin almost completely suppressed LPS-mediated upregulation of nuclear β-catenin (Figure 6D). Further results showed that melatonin decreased β-catenin both in cytoplasm (Figure 6E) and nucleus (Figure 6F) in LPS-unstimulated DU145 cells in a dose-dependent manner. Next, we investigated the effect of melatonin on phosphorylation of GSK-3β on serine 9 in the “destruction complex” by co-immunoprecipitation. As shown in Figure 6G, p-GSK3βSer9 in the “destruction complex” was upregulated in LPS-stimulated DU145 cells. Melatonin inhibited LPS-mediated upregulation of p-GSK3βSer9. Interestingly, melatonin suppressed the phosphorylation of GSK3βSer9 in the “destruction complex” in LPS-unstimulated DU145 cells (Figure 6G). In addition, β-catenin ubiquitination was downregulated in LPS-stimulated DU145 cells (Figure 6H). Melatonin reversed LPS-mediated downregulation of β-catenin ubiquitination. Of interest, melatonin promoted β-catenin ubiquitination in LPS-unstimulated DU145 cells (Figure 6H).

Figure 6

Melatonin inhibits β-catenin nuclear translocation in LPS-stimulated and -unstimulated DU145 cells. (A and B) DU145 cells were incubated with LPS (2 μg/mL) for indicated time. (A) Phosphorylated β-catenin in cytoplasm was measured. (B) Nuclear β-catenin was measured. (C) DU145 cells were incubated with LPS (2 μg/mL) with/without melatonin (2 mM) for 2 h. Phosphorylated β-catenin in cytoplasm was measured. (D) DU145 cells were incubated with LPS (2 μg/mL) in absence or presence of melatonin (2 mM) for 6 h. Nuclear β-catenin was measured. (E and F) DU145 cells were incubated with melatonin for indicated dose. The levels of β-catenin in (E) cytoplasm and (F) nuclear were measured. (G and H) DU145 cells were incubated with LPS (2 μg/mL) in absence or presence of melatonin (2 mM) for 2 h. (G) p-GSK3βSer9 in “destruction complex” was determined using Co-IP. (H) Ubiquitination of cytoplasm β-catenin was determined using Co-IP. All experiments were repeated three times. Data were expressed as means ± S.E.M., *P<0.05, **P<0.01 versus control group, ##P<0.01 versus LPS group.

Discussion

This study investigated the effects of LPS on PCa cells migration and invasion, and evaluated the protective effects of melatonin. The results showed that LPS enhanced the migration and invasion of PCa cells. In addition, LPS stimulated inflammatory reaction and induced epithelial-mesenchymal transformation (EMT) of PCa cells by activating TLR4 downstream IL-6/STAT3, AKT/GSK-3β and β-catenin pathways. Melatonin suppressed LPS-mediated migration and invasion in PCa cells. Of interest, melatonin also inhibited migration and invasion in LPS-unstimulated PCa cells. Mechanistically, melatonin inhibited the EMT of PCa cells through down-regulating IL-6/STAT3, AKT/GSK-3β and β-catenin pathways. This study provides evidence that melatonin inhibits migration and invasion of PCa cells through blocking multiple TLR4 downstream EMT-associated pathways in LPS-stimulated and -unstimulated PCa cells.

Epithelial–mesenchymal transition (EMT) is an important process of tumor metastasis. Cancer cells acquire aggressive phenotypes by activating multiple signals and downstream transcription factors.12,22 In this study, our result showed that the epithelial marker, E-cadherin, was downregulated in LPS-treated PCa cells. By contrast, two mesenchymal markers, vimentin and N-cadherin, were upregulated in LPS-stimulated PCa cells. Moreover, melatonin induced E-cadherin expression and decreased the levels of vimentin and N-cadherin in LPS-stimulated and -unstimulated PCa cells. Moreover, the nuclear levels of four key EMT-related transcription factors Zeb1, Snail, Slug and Twist were all elevated in LPS-treated PCa cells. Melatonin downregulated nuclear levels of Zeb1, Snail, Slug and Twist in LPS-stimulated and -unstimulated PCa cells. These results indicated that LPS induced migration and invasion by promoting EMT in PCa cells. Melatonin suppressed EMT in LPS-stimulated and -unstimulated PCa cells.

NF-κB is a key signal molecule in response to LPS/TLR4-mediated inflammatory responses.23,24 Once LPS binds to Toll-like receptor 4 (TLR4), TLR4 undergoes conformational changes to recruit some downstream proteins, including MyD88 and TIRAP, which initiates a complex signal transduction cascade, eventually lead to the activation of NF-κB and increased transcription of different cytokines.25 Our results showed that LPS activated NF-κB and up-regulated the expressions of pro-inflammatory cytokines and chemokines in PCa cells. Previous study indicated that melatonin binding to MT1 receptor triggers Gαq and Gαs proteins activation and leads to an increase of cyclic AMP (cAMP). Increased intracellular cAMP then activates protein kinase A and C, which inhibit NF-κB constitutive activation in PCa cells.26 As expected, melatonin inhibited NF-κB activation and attenuated LPS-induced upregulation of chemokines and pro-inflammatory cytokines. In addition, melatonin inhibited expressions of pro-inflammatory cytokines and chemokines in LPS-unstimulated PCa cells. These results suggested that melatonin suppressed NF-κB mediated inflammatory signaling activation not only in LPS-stimulated but also in LPS-unstimulated PCa cells.

IL-6/STAT3, a key TLR4/NF-κB downstream signal, initiates EMT by inducing expression of EMT-relative transcription factors, such as Zeb1, Snail and Twist1.11,27 In this study, our results provided evidence that LPS activated IL-6/STAT3 signal in PCa cells. Melatonin inhibited IL-6/STAT3 activation in LPS-stimulated PCa cells. In addition, melatonin inhibited IL-6/STAT3 signal activation in LPS-unstimulated PCa cells. These results suggested that melatonin inhibits migration and invasion of PCa cells, at least partially, by suppressing IL-6/STAT3-mediated EMT in LPS-stimulated and -unstimulated PCa cells.

Previous studies have demonstrated that AKT/GSK-3β is involved in LPS/TLR4-mediated inflammatory response.28,29 In this study, our results are similar; LPS induced AKT phosphorylation at serine 473 (p-AKTSer473) in DU145 cells. GSK-3β is a downstream target of AKT, constitutively activated GSK-3β is inactivated by the activated AKT through phosphorylation at serine 9.30 As expected, the level of phosphorylated GSK3β at serine 9 (p-GSK3βSer9) was elevated in LPS-stimulated DU145 cells. A report from our laboratory found that melatonin inhibited TLR4-mediated Akt phosphorylation in LPS-induced macrophages.31 In this study, our results showed that melatonin inhibited LPS-mediated elevation of p-AKTSer473 and p-GSK3βSer9. In addition, melatonin downregulated the levels of p-AKTSer473 and p-GSK3βSer9 in LPS-unstimulated DU145 cells. Increasing evidence demonstrated that AKT/GSK3β signal induced EMT via stabilizing snail.32,33 These results suggested that melatonin inhibits migration and invasion of PCa cells, at least in part, by suppressing AKT/GSK-3β-mediated EMT.

Accumulating data has shown that β-catenin is another key regulator for cancer-relative EMT.34,35 As a core component of the canonical Wnt signaling, β-catenin is degraded by the ubiquitin-proteasome system, due to phosphorylation by a “destruction complex” composed with Glycogen synthase kinase 3β (GSK3β)-Adenomatous polyposis coli (APC)-Axin-Casein kinase 1(CK1).36,37 Blocking “destruction complex” activity leads to the cytoplasmic accumulation of β-catenin which in turn causes the transnucleation of β-catenin, and triggers target gene transcription by replacing transcription repressors and binding to TCF/LEF.37 Some of these target genes are EMT-relative transcription factors such as Zeb1 and Twist.38,39 Previous studies indicated that LPS stimulated cytosolic β-catenin accumulation and induced macrophage migration by inhibiting GSK3β.40 In this study, our result showed that nuclear β-catenin was upregulated in LPS-stimulated DU145 cells. Melatonin inhibited LPS-mediated upregulation of nuclear β-catenin in DU145 cells. Interestingly, melatonin downregulated β-catenin levels both in cytoplasm and nuclear in LPS-unstimulated DU145 cells. GSK-3β, as a key component of the “disruption complex”, can recognize and phosphorylate β-catenin to generate the binding site of β-TrCP, a key component of E3 ubiquitin ligase, which is then primes β-catenin ubiquitinated and degraded.36 In this study, our results showed that the level of p-GSK-3βSer9 in “destruction complex” was increased in LPS-treated DU145 cells. Correspondingly, phosphorylation and ubiquitination levels of cytoplasmic β-catenin were decreased in LPS-treated DU145 cells. Melatonin inhibited GSK-3βSer9 phosphorylation of “destruction complex” in DU145 cells and reversed LPS-mediated downregulation of phosphorylation and ubiquitination of cytoplasmic β-catenin in DU145 cells. Moreover, melatonin promoted β-catenin phosphorylation and ubiquitination in LPS-unstimulated DU145 cells. These results indicated that melatonin inhibited β-catenin nuclear translocation through reinforcing phosphorylation and ubiquitination degradation of cytoplasmic β-catenin not only in LPS-stimulated but also in LPS-unstimulated PCa cells.

Conclusion

In summary, the present study investigated the effects of LPS on PCa cells migration and invasion, and evaluated the protective effects of melatonin. Our results indicated that LPS promoted migration and invasion by activating the EMT in PCa cells. Mechanistically, LPS induced the EMT of PCa cells through at least three different TLR4 downstream pathways, including NF-κB/IL-6/STAT3, AKT/GSK-3β and β-catenin. Melatonin suppressed migration and invasion by blocking EMT mediated by the above three pathways both in LPS-stimulated and -unstimulated PCa cells. Taken together, this study demonstrates the potential impact of LPS on PCa cell migration and invasion. This study not only sheds new light into understanding the role of bacterial infection in PCa metastasis but also provides a potential therapeutic agent. Further in vivo studies are needed to verify these results.