Background
Persistent inflammatory conditions can induce tumorigenesis [
1]. Hepatocellular carcinoma (HCC) is closely associated with chronic inflammatory liver diseases and the endotoxin level in the portal and peripheral veins of those patients is higher owing to changes in the intestinal mucosal permeability and increased bacterial infection [
2‐
4]. Lipopolysaccharide (LPS), a cell wall constituent of gram-negative bacteria, is released during lysis of bacteria. It has been reported that LPS can induce cytokines from immune cells and inhibit tumor growth [
5,
6], but recent studies have shown that LPS can alter cytokine levels in the tumor microenvironment and exert direct effects on tumor cell proliferation, invasion and metastasis
in vitro and
in vivo [
7‐
10].
Toll-like receptor 4 (TLR4), the receptor for LPS, is not only important in the regulation of immune responses to infection [
11], but also is involved in noninfectious inflammatory diseases, such as tumor invasion and survival [
12]. TLR4 has been detected in many human cancer cell lines, including pancreatic, lung, breast, prostate, liver and colorectal cancer [
10,
12‐
16]. When silencing TLR4 expression, the invasion, survival, and tumorigenicity of human prostate cancer cells was inhibited, which indicates TLR4 plays a significant role in connecting inflammation and cancer invasion and progression [
12]; however, the exact mechanism is still not clear.
Epithelial to mesenchymal transition (EMT) is a process in which epithelial cell layers lose polarity together with cell to cell contacts which results in a dramatic remodeling of the cytoskeleton, and has an important role in tumor metastasis [
17]. When human intrahepatic biliary epithelial cells (IBECs) were exposed to high levels of LPS, the IBECs could undergo EMT, potentially contributing to hepatic fibrosis or even hepatoma [
18]. In HCC tissues, LPS was previously reported to promote adhesion and invasion in hepatoma cells [
9]. These effects suggest that LPS-induced TLR4 signaling provides a survival benefit for metastatic tumors; however, whether TLR4 signaling can induce EMT in HCC cells and the mechanism involved remains unclear.
In the present study, we provide evidence that LPS-induced TLR4 signaling promotes HCC cell invasion and EMT in vitro and in vivo, and a high expression of TLR4 in HCC tissues was strongly associated with both poor cancer-free survival and overall survival in patients, which indicates that LPS is closely related to tumor invasion and metastasis, rather than only anti-tumor effects.
Methods
Reagents and antibodies
(Dulbecco's) modified Eagle's medium ((D)MEM), fetal bovine serum (FBS), penicillin, streptomycin sulfate, glutamine, and 0.05% trypsin/0.02% ethylenediamine tetraacetic acid (EDTA) solution were purchased from Invitrogen (Carlsbad, CA, USA). LPS derived from Escherichia coli strain 055:B5, Trizol, Lipofectamine 2000 and BAY11-7028, the inhibitor of nuclear factor kappa B (NF-κB), were purchased from Sigma (St. Louis, MO, USA). A dual luciferase reporter gene assay kit to detect NF-κB activity was purchased from Promega (Madison, WI, USA). Rabbit anti-human E-cadherin and Vimentin antibodies were obtained from Thermo (Fremont, CA, USA). Goat anti-human Snail antibody was purchased from R&D (Minneapolis, MN, USA). Rabbit anti-human TLR4 antibody was purchased from Bioworld Technology (St. Louis, MN, USA).
Cell lines and culture condition
HCC cell lines (HepG2, Huh7, Hep3B, SMMC-7721 and MHCC97-H) were cultured in (D)MEM supplemented with 10% FBS, 100 U/ml penicillin, 100 μg/ml streptomycin sulfate, and 2.0 mM glutamine. Cells were maintained at 37°C in a humidified 5% CO2 atmosphere and subcultured by trypsinization with 0.05% trypsin/0.02% EDTA when cells became confluent. For LPS treatment, subconfluent cultures of cells were treated with 10 μg/ml LPS for 48 hours. For NF-κB inhibition, 10 μM BAY11-7028 was added to the cell culture at the same time as LPS.
Patients and tissue specimens
Specimens of HCC tissues were obtained from 106 HCC patients who underwent hepatic resection at the Eastern Hepatobiliary Surgery Hospital of the Second Military Medical University from October 2000 to November 2003. These patients included 88 men and 18 women with a median age of 49 years (range: 11 to 72), and all of the specimens were subjected to immunohistochemisty (IHC). Prior informed consent was obtained and the study protocol was approved by the Ethics Committee of the Eastern Hepatobiliary Surgery Hospital.
Wound healing and Transwell assay
The methods for wound healing and the Transwell assay have been described [
19‐
21]. For the wound-healing assay, cells (5 × 10
4) were seeded on a 24-well dish and incubated for 24 hours, the monolayer was then disrupted with a cell scraper (1.2 mm width), and photographs were taken at 0 and 48 hours in a phase-contrast microscope. Experiments were carried out in triplicate, and four fields of each point were recorded. For the Transwell assay, Boyden chambers (8 μm pore size) were coated with 200 μl Matrigel at 200 μg/ml and incubated overnight. Cells (5 × 10
4) in medium without serum were plated in the upper chamber, and the medium containing 5% FBS was added in the lower chamber as a chemoattractant. After 24 hours of incubation at 37°C, the cells were fixed in 4% formaldehyde and stained with crystal violet dye, and the cells that invaded through the pores to the lower surface of the filter were counted under a microscope. Three invasion chambers were used per condition. The values obtained were calculated by averaging the total number of cells from three filters.
All procedures involving animals were performed in accordance with the institutional animal welfare guidelines of the Second Military Medical University. Cells were injected into the splenic vein of eight-week-old nude mice (BALB/c strain) at 1 × 106 cells/injection site. The mice were sacrificed after six weeks and the numbers of surface liver metastases were counted.
Quantitative real-time polymerase chain reaction (qPCR)
Total RNA extraction, complementary DNA (cDNA) synthesis, and qPCR were performed as described [
22]. The primer sequences used in qPCR are shown in Additional file
1 Table S1.
Western-blot analysis
Total soluble protein extractions from cultivated cells and western-blot analysis were performed as described [
22]. Antibodies used in western-blot experiments were specific for TLR4, Snail or E-cadherin.
Immunofluorescence
About 104 cells were seeded on a 48-well dish. After 24 hours, the cells were washed twice with PBS, and fixed in 4% paraformaldehyde and 0.1% Triton × 100 in PBS buffer at 4°C for 30 minutes. After being washed with PBS, the cells were incubated with the blocking solution (10% goat serum in PBS), and then incubated overnight with the primary antibodies, washed with PBS, and finally incubated with secondary antibodies at 37°C for 2 hours. After being stained with DAPI/PI (4', 6-diamidino-2-phenylindole/ propidium iodide), all matched samples were photographed using an immunofluorescence microscope and identical exposure times.
FACS Analysis
The expression of TLR4 on HCC cells was determined using indirect immunofluorescent staining. In this analysis, 20 μl of anti-TLR4 antibody were added to 100 μl of cell suspension (1 × 106 cells/ml) and incubated at 4°C for 30 minutes, and further stained with secondary fluorescein isothiocyanate (FITC)-conjugated antibody at 4°C for 30 minutes. FITC-conjugated isotype IgG antibody was used as a negative control. The stained cells were analyzed on a FACS flow cytometer (Becton Dickinson, San Jose, CA, USA).
Generation of stable TLR4-expressed and knocked-down cell lines
The adeno-associated viral construct which directs the expression of TLR4 (pAAV-TLR4-IRES-GFP), siRNA against TLR4 (pAAV-siTLR4-IRES-GFP) and their control vectors were constructed and verified by DNA sequencing. The adeno-associated viral construct that expresses TLR4 or siRNA targeting TLR4 were co-transfected with adeno-associated viral packaging plasmids pAAV-RC, pAAV-helper into 293T cells by Lipofectamine 2000 according to the manufacturer's instructions. At 24 and 48 hours post-transfection, culture medium was collected to be incubated with target cells. At 48 hours post-infection, infected cells were harvested for gene and protein expression analysis or selected with G418 (400 μg/ml) for two weeks to establish stable clones.
NF-κB luciferase reporter assays
NF-κB luciferase reporter assays were performed as described previously [
23]. Briefly, cells were co-transfected with a pNF-κB-luc reporter construct and a renilla luciferase-expressing plasmid (internal control to normalize for transfection efficiency) using Lipofectamine 2000 according to the manufacturer's instructions. After treatment for 48 hours, firefly and renilla luciferase activities were assessed using a dual luciferase reporter gene assay kit. NF-κB transcriptional activity = relative light units of firefly luciferase/relative light units of renilla luciferase.
IHC analysis
The sample processing and IHC procedures were performed as described [
20] and the results were expressed as the percentage of tumor cells with a positive stain. Thereafter, the percentage of TLR4-positive tumor cells was scored on a scale of 0 to 4 (0: no staining; 1: ≤10%; 2: 11% to 30%; 3: 31% to 50%; 4: ≥50%). Furthermore, the expression level of TLR4 was divided into the following two groups according to score: low (0, 1, 2) and high (3, 4) [
16].
Statistical analysis
All data, expressed as mean ± the standard error of the mean (SEM), were from at least three separate experiments. Groups were compared by analysis of variance (ANOVA) with a posteriori contrast by least significant difference; or by Student t-test using the Microsoft Excel Analysis ToolPak (Microsoft, Redmond, WA). P < 0.05 was considered significant.
Discussion
To study the key roles of TLR4 in LPS-induced EMT of HCC cells, we selected four different HCC cells (HepG2, Huh7, Hep3B, and SMMC-7721) and carefully evaluated the direct effect of LPS on the invasive potential of these cells. We demonstrated that LPS could significantly enhance the invasive potential and induce EMT of HCC cells with higher expression of TLR4, which was confirmed to be in a TLR4 dependent manner. Further studies showed that LPS could directly activate NF-κB signaling through TLR4 in HCC cells. Importantly, blocking NF-κB signaling significantly inhibited Snail expression and inhibited EMT occurrence. Finally, we revealed that high expression of TLR4 in HCC tissues was strongly associated with both poor cancer-free survival and overall survival. Our results indicate that TLR4 is necessary for LPS-induced EMT and metastasis in HCC cells, which increases our understanding of the pathogenesis and provides clues for developing new strategies against LPS-related metastasis of HCC.
This study provides the first evidence that LPS-induced metastasis and EMT of HCC cells is TLR4 dependent. EMT is a key event in the metastasis of tumors from epithelial cells. An imaging study
in vivo has shown that carcinoma cells migrate from mouse primary tumors through a process of EMT and that this process is dependent on an inflammatory microenvironment [
26]. Recently, the new finding that TNF-α induces EMT in MCF-7 breast cancer cells reinforced the connection between inflammation and EMT [
27]. In HCC patients, the level of endotoxin in the portal and peripheral veins has been determined to be higher, which results in a high level of LPS in the portal vein. In this study, four different human HCC cell lines (HepG2, Huh7, Hep3B, and SMMC-7721) were tested in the scratch wound-healing and Matrigel invasion assay. Our results showed that not all tumor cell lines responded
in vitro to LPS, the motility and invasion abilities of SMMC-7721 cells were enhanced significantly, and the data of the splenic vein metastasis assay
in vivo were consistent with this. However, Liu
et al. have inferred that LPS may be ignored by immune systems and may directly promote the invasion abilities of HepG2 and HepG2.2.15 cells [
9]. Our experiment showed that LPS had little effect on HepG2 cells; the different experimental methods may have led to the inconsistent results.
Administration of LPS can induce EMT of human IBECs, which makes tumor cells undergo a phenotypic change to produce bipolar cells with a fibroblastic morphology. In these four HCC cells, LPS induced EMT in SMMC-7721 cells but not the others, and SMMC-7721 cells constitutively expressed a high level of TLR4, which might be associated with LPS-induced metastasis and EMT of HCC cells. When we used adeno-associated virus to express TLR4 in HepG2 cells or knocked down TLR4 in SMMC-7721 cells with siRNA, the metastasis and EMT of HCC cells were changed by being treated with LPS. All of the above results highlight a direct link between LPS and tumor cell metastasis. The effect of LPS is possible because SMMC-7721 cells constitutively express TLR-4. Not all tumor cells are positive for TLR-4, and this could depend on the tissue from which the tumor cells originated. Molteni
et al. analyzed the responses of tumor cell lines from different origins (melanoma, ovarian carcinoma, neuroblastoma) to LPS
in vitro; the results showed that only melanoma cells significantly increased cell adhesion, when triggered with LPS and these effects were associated with the constitutive expression of TLR-4 [
28]. Different HCC cell lines also have different expression of TLR4, and LPS-induced EMT in cells is in a TLR4 dependent manner as well. However, notably, HCC development is a multifactorial and complicated process, which has a close association with various risk factors, such as viral infection, alcohol consumption, steatosis, toxic substances and inflammation. In this process, LPS is an important EMT inducer but not the only one. Many gene alterations and cytokines also could induce EMT and promote metastasis [
29]. Therefore, although HCC cells with low expression or even a lack of TLR4 are not susceptible to LPS, they might perform EMT induced by other TLR4-independent mechanisms.
NF-κB activation is well known to play an important role in responses to LPS via TLR-4 [
8,
24]. Previous studies have shown that LPS may be able to induce NF-κB activation in colon and pancreatic cancer cells [
30,
31], which strongly suggests that LPS can act not only on immune cells but also on some cancer cells from epithelial cells. In the current study, we found that ligation of LPS and TLR4 could activate the NF-κB pathway in SMMC-7721 cells, but not in HepG2 cells. Furthermore, activation of NF-κB is involved in the invasion and metastasis of HCC. Many studies have shown that NF-κB is a key regulator of Snail expression in cancer cell lines and metastatic tumor samples [
32‐
35]. In our study, activation of NF-κB in SMMC-7721 cells up-regulated Snail expression, which was a prerequisite for these cells to undergo an EMT toward an invasive, metastatic tumor phenotype. Snail is identified as a transcription factor in the control of EMT, and expression of Snail represses expression of E-cadherin and induces EMT in breast cancer cells, indicating that Snail plays a fundamental role in EMT [
36,
37]. To confirm the contribution of NF-κB activation to the enhanced invasive ability and EMT, we used BAY-11-7082 to inhibit NF-κB activation in SMMC-7721 cells, and the inhibitor not only reduced NF-κB activation, but also down-regulated Snail expression. Our data strongly suggest that activation of the NF-κB pathway by LPS is TLR4-dependent and increases Snail expression to induce EMT.
TLR4 expression has been described in different human tumors [
10,
12‐
16]. Gonzalez-Reyes
et al. have showed that TLR4 expression in breast carcinomas was associated with an increased incidence of metastasis and has prognostic significance [
14]. Cammarota
et al. analyzed 116 tissue samples from patients with different stages of colorectal disease and found that adenocarcinoma patients with higher TLR-4 expression in the stromal compartment had a significantly increased risk of disease progression and relapsed significantly earlier than those with lower expression levels [
38]. Interestingly, analysis of the association of TLR4 expression and the clinicopathological characteristics of 106 HCC patients reveals that TLR4 expression is significantly correlated with margin, vascular invasion, portal vein thrombosis, and so on, which are widely accepted markers for metastasis and poor prognosis of HCC [
39,
40]. The Kaplan-Meier analysis shows that the HCC patients with high TLR4 expression in general had a shorter cancer-free interval and a worse overall survival than those with low expression, suggesting that TLR4 may be a useful biomarker of HCC metastasis and recurrence.
In conclusion, our study has shown for the first time that LPS-induced EMT and the invasive potential of HCC cells is in a TLR4 dependent manner and that activation of the NF-κB-Snail regulatory axis may function as a therapeutic target. Furthermore, we have demonstrated that high expression of TLR4 in HCC tissues significantly correlates with metastasis and recurrence of HCC. Our data suggest TLR4 as a novel prognostic marker and a potential therapeutic target for LPS-induced EMT and metastasis of HCC.
Acknowledgements
This project was supported by Key project of National Natural Science Foundation of China (Grant NO: 81030041); Key Basic Research Project of China (Grant NO: 2010CB945600, 2011CB966200); National Natural Science Foundation of China (Grant NO: 30901722, 31171321, 81000970, 81101622); Special Funds for National Key Sci-Tech Special Project of China (Grant NO: 2008ZX10002-019, 2008ZX10002-025); Shanghai Science and Technology Committee (Grant NO: 10ZR1439600, 11ZR1449500, 10411963100, 10ZR1439900); Science Fund for Creative Research Groups, NSFC, China (Grant NO: 30921006); Stem Cell and Medicine Research Center's Innovation Research Program (Grant NO: SCMRC1102); 2011 Ph.D. Graduate Innovation Fund Project (Grant NO: 27).
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
YYJ, ZPH, KS and LXW participated in the design and performance of the study. SSZ, JH and YL provided clinical data and performed statistical analysis. RL, LG, XZ and QDZ carried out cell culture, molecular genetic studies and the mouse experiment. MCW and LXW conceived of the study and participated in its design and coordination. The manuscript was drafted by YYJ, ZPH and KS, and reviewed by all authors. All authors approved the final version of the manuscript to be published.