Thrombin is a therapeutic target for metastatic osteopontin-positive hepatocellular carcinoma


  • Yu-Hua Xue,

    1. Liver Cancer Institute and Zhongshan Hospital, Institutes of Biomedical Sciences, Fudan University, Shanghai, China; Key Laboratory of Carcinogenesis and Cancer Invasion, Ministry of Education, China
    Search for more papers by this author
  • Xiao-Fei Zhang,

    1. Liver Cancer Institute and Zhongshan Hospital, Institutes of Biomedical Sciences, Fudan University, Shanghai, China; Key Laboratory of Carcinogenesis and Cancer Invasion, Ministry of Education, China
    Search for more papers by this author
  • Qiong-Zhu Dong,

    1. Liver Cancer Institute and Zhongshan Hospital, Institutes of Biomedical Sciences, Fudan University, Shanghai, China; Key Laboratory of Carcinogenesis and Cancer Invasion, Ministry of Education, China
    Search for more papers by this author
  • Jian Sun,

    1. Liver Cancer Institute and Zhongshan Hospital, Institutes of Biomedical Sciences, Fudan University, Shanghai, China; Key Laboratory of Carcinogenesis and Cancer Invasion, Ministry of Education, China
    Search for more papers by this author
  • Chun Dai,

    1. Liver Cancer Institute and Zhongshan Hospital, Institutes of Biomedical Sciences, Fudan University, Shanghai, China; Key Laboratory of Carcinogenesis and Cancer Invasion, Ministry of Education, China
    Search for more papers by this author
  • Hai-Jun Zhou,

    1. Liver Cancer Institute and Zhongshan Hospital, Institutes of Biomedical Sciences, Fudan University, Shanghai, China; Key Laboratory of Carcinogenesis and Cancer Invasion, Ministry of Education, China
    Search for more papers by this author
  • Ning Ren,

    1. Liver Cancer Institute and Zhongshan Hospital, Institutes of Biomedical Sciences, Fudan University, Shanghai, China; Key Laboratory of Carcinogenesis and Cancer Invasion, Ministry of Education, China
    Search for more papers by this author
  • Hu-Liang Jia,

    1. Liver Cancer Institute and Zhongshan Hospital, Institutes of Biomedical Sciences, Fudan University, Shanghai, China; Key Laboratory of Carcinogenesis and Cancer Invasion, Ministry of Education, China
    Search for more papers by this author
  • Qin-Hai Ye,

    Corresponding author
    1. Liver Cancer Institute and Zhongshan Hospital, Institutes of Biomedical Sciences, Fudan University, Shanghai, China; Key Laboratory of Carcinogenesis and Cancer Invasion, Ministry of Education, China
    • Liver Cancer Institute and Zhongshan Hospital, Fudan University, 180 Feng Lin Road, Shanghai 200032, China
    Search for more papers by this author
  • Lun-Xiu Qin

    Corresponding author
    1. Liver Cancer Institute and Zhongshan Hospital, Institutes of Biomedical Sciences, Fudan University, Shanghai, China; Key Laboratory of Carcinogenesis and Cancer Invasion, Ministry of Education, China
    • Liver Cancer Institute and Zhongshan Hospital, Fudan University, 180 Feng Lin Road, Shanghai 200032, China
    Search for more papers by this author
    • fax: +86-21-5423 7960

  • Potential conflict of interest: Nothing to report.


We previously identified osteopontin (OPN) as a promoter and thus a potential therapeutic target for hepatocellular carcinoma (HCC) metastasis. The serine protease thrombin interacts with OPN and can modify its biological activity. To explore the role of thrombin alone or in conjunction with OPN in HCC, we studied the correlation of thrombin levels to HCC prognosis in patients with various OPN levels, and evaluated the effects of OPN fragments generated by thrombin cleavage on proliferation and adhesion of HCC cells. We found that the thrombin level was strongly associated with the metastatic potential of HCC cell lines, and that thrombin was remarkably overexpressed in HCC tissue compared with adjacent nontumor tissue. In addition, HCC tissue from patients with recurrent disease displayed much higher thrombin levels, particularly in those with elevated OPN levels. Only HCCs with elevated OPN levels had a significant correlation between high thrombin levels and overall survival (OS; P < 0.01), or time to recurrence (TTR; P < 0.0001) of HCC. Multivariate analysis revealed that thrombin was an independent prognostic indicator. In vitro assays demonstrated that thrombin promotes the proliferation and adhesion of OPN+ HCC cells. Furthermore, thrombin activated the focal adhesion kinase (FAK) pathway of OPN+ HCC cells, which was blocked by the inhibition of integrin β1. Conclusion: Thrombin plays an important role in OPN-mediated aggressive phenotype of HCC through activation of integrin β1-FAK signaling, and is an independent poor prognostic factor for HCC. Thus, thrombin may be a potential therapeutic target to inhibit HCC metastasis in OPN+ patients (HEPATOLOGY 2010.)

Osteopontin (OPN) is an extracellular matrix (ECM) protein that binds to αvβ integrins and receptors of the CD44 family to propagate cellular signals and promotes induction of cell adhesion, chemotaxis, ECM degradation, angiogenesis, prevention of apoptosis, and indolent tumor growth.1,2 Many studies have shown that increased OPN levels are associated with increased aggressiveness and metastatic potential of hepatocellular carcinoma (HCC) and are positively correlated with poor prognosis and early tumor recurrence in patients with HCC.3-5 Thus, the molecules involved in the signaling pathways through which OPN mediates cancer metastasis, especially the portion of the pathway mediating the early stages of cellular invasion, may contain potential therapeutic targets for HCC metastasis.6

Thrombin is a serine protease that performs a multifaceted role in coagulation. Thrombin cleaves OPN at the cleavage site (RSK) into two fragments of approximately equivalent size, which changes the topological structure of OPN to display the integrin and CD44 binding domains.7 This cleavage by thrombin improves the bioactivity of OPN and is necessary for efficient engagement with the integrin receptor.8-11 Previous studies have demonstrated that thrombin-cleaved OPN is critically involved in the pathogenesis of various diseases.12-14 Thrombin has also been shown to contribute to tumor progression in manners both coagulation-dependent and coagulation-independent.15, 16 However, the possible mechanism for how thrombin and OPN are involved in HCC metastasis is not yet known. Furthermore, the prognostic value of thrombin and OPN for HCC has not yet been established.

In this study we investigated the messenger RNA (mRNA) and protein levels of thrombin and OPN in HCC cell lines with various metastatic potential and in clinical HCC samples, evaluating the effects of thrombin treatment on both in vitro adhesion and proliferation abilities of HCC cells with relatively high OPN expression (PLC-OPN). We also explored the possible mechanisms involved in the effects of thrombin and OPN on HCC metastasis.


cDNA, complementary DNA; Ct, cycle threshold; FAK, focal adhesion kinase; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; HCC, hepatocellular carcinoma; mRNA, messenger RNA; OPN, osteopontin; OS, overall survival; qRT-PCR, quantitative real-time polymerase chain reaction; TBP, TATA-binding protein; TNM, tumor-node-metastasis; TTR, time to recurrence.

Patients and Methods

Patients and Follow-up.

A total of 302 patients were enrolled in this study and underwent curative liver resection for HCC at the Liver Cancer Institute and Zhongshan Hospital, Fudan University (Shanghai, China) between 2003 and 2006. For each patient complete follow-up data were available and the diagnosis of HCC was confirmed by two pathologists.

Tissue specimens obtained from 230 consecutive HCC patients with well-preserved liver function (Child-Pugh A class) who underwent curative resection without preoperative treatment were used in immunohistochemistry studies. The detailed clinicopathological characteristics are summarized in Table 1. Frozen tissue samples were collected from 72 HCC patients who had primary HCC with a solitary tumor and without major vascular invasion or regional lymph node or distant extrahepatic metastasis and used in real-time polymerase chain reaction (PCR) studies and western blot analyses. The noncancerous hepatic tissues were dissected 2 to 5 cm away from the tumor. The detailed clinicopathological characteristics of these 72 patients are summarized in Supporting Information Table S1. Twenty normal liver tissues were collected from patients with liver hemangioma and used as controls. Tissue samples were collected immediately after resection, transported in liquid nitrogen, and stored at −80°C until use.

Table 1. Association of Thrombin Expression Detected by Immunuohistochemical Staining with the Clinicopathologic Characteristics of HCC Patients with Various Osteopontin Expressions
 Whole Study GroupOsteopontin–Positive GroupOsteopontin–Negative Group
Thrombin ExpressionThrombin ExpressionThrombin Expression
Negative (n = 153)Positive (n = 77)PNegative (n = 54)Positive (n = 36)PNegative (n = 99)Positive (n = 41)P
  • ALT, alanine aminotransferase; TNM, tumor-node-metastasis; HBsAg, hepatitis B surface antigen.

  • *

    Fisher's exact tests, and χ2 tests for all other analyses.

 Male12964 4731 8233 
 >527634 3117 4517 
Preoperative serum AFP (ng/mL)
 >208746 2525 6221 
 Positive13064 4329 8735 
Liver cirrhosis
 Yes14368 5231 9137 
ALT (units/L)
 >75196 72 124 
Tumor size (cm)
 >56242 1723 4519 
Tumor number
 Multiple2411 105 146 
Vascular invasion
 Yes2220 513 177 
TNM stage
 II2519 914 165 
 III1615 56 119 
Tumor differentiation
 III-IV3825 1417 248 
Tumor encapsulation
 None3826 2323 153 

Clinical samples were collected from these patients after obtaining informed consent according to an established protocol approved by the Ethics Committee of Fudan University (Shanghai, China). The data did not contain any information that could lead to patient identification.

All patients received follow-up care until March 15, 2009. Patients were monitored every 2 months postsurgery as described.17 A diagnosis of recurrence was based on typical appearance on computed tomography (CT) and/or magnetic resonance imaging (MRI). The recurrent tumors were treated as described in previous studies.18

The endpoints included the time to recurrence (TTR) and overall survival (OS). TTR was defined as the time between the surgery and the first report of intrahepatic or distant recurrence (excluding patients who died before recurrence from causes unrelated to HCC). TTR was recorded as the date of death or of last follow-up for patients who had not experienced a recurrence at the time of death or last follow-up, respectively. OS was defined as the interval between the dates of surgery and death.19

Quantitative Real-Time PCR.

Total RNA was extracted from cell lines and frozen tumor specimens using Trizol Reagent (Invitrogen, Carlsbad, CA). Total RNA (5 μg) was reverse transcribed using oligo dT and SuperScript III reverse transcriptase according to the manufacturer's instructions. The complementary DNA (cDNA) was diluted 1:50 in water and 4.5 μL of this mixture was used as template in a 10 μL quantitative PCR (qPCR) reaction. Amplification and detection were performed using the ABI PRISM 7900 Sequence Detection System (Applied Biosystems, Foster City, CA). Briefly, the cycle conditions were as follows: 50°C for 2 minutes (required for optimal AmpErase UNG activity), template denaturation at 95°C for 10 minutes, 40 cycles of denaturation at 95°C for 15 seconds, and combined primer annealing/elongation at 60°C for 1 minute. In addition, qPCR of TATA-binding protein (TBP) was used as an endogenous control to normalize for differences in the amount of total RNA in each sample. Relative expression of genes was calculated and expressed as 2-&Dgr;Ct (cycle threshold) as described.20 The following probes were used for detection: Hs00959010_m1 for OPN, Hs00354679_m1 for thrombin, Hs00236976_m1 for integrin-β1, and Hs00427620_m1 for TBP.

Immunohistochemical Staining.

Formalin-fixed and paraffin-embedded tissues (both tumor and nontumor liver tissues) were used for immunohistochemical staining. Following deparaffinization, 4-μm tissue sections were rehydrated and subjected to antigen retrieval by microwaving in 0.01 mol/L sodium citrate (pH 6) for 10 minutes. Sections were stained with monoclonal antimouse OPN antibody and thrombin antibody (Santa Cruz Biologicals, Santa Cruz, CA), using a two-step immunoperoxidase technique performed as described.18 Briefly, after microwave antigen retrieval tissue sections were incubated with primary antibodies for 60 minutes at room temperature. Following 30 minutes of incubation with secondary antibody, the sections were developed in diaminobenzidine solution under microscopic observation and counterstained with hematoxylin. Negative controls were stained identically, but without primary antibody incubation. The intensity of positive staining was measured as described in the Supporting Information. Based on the intensity of thrombin or OPN in the central positive staining area in tumor section as a cutoff value, the intensity of thrombin or OPN was classified positive (thrombin+ ≥20%, and OPN+ ≥5% of tumor section) and negative (thrombin− <20%, and OPN− <5%).

Detection of Protein by Western Blot.

Conditioned media and cell lysate were prepared as described.21 The protein expression levels of thrombin, OPN, FAK, Phospho-FAK (Tyr397), and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were evaluated by western blot. Twenty to 50 μg of total protein extracted from the tissues or cells was separated by way of sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred onto polyvinylidene fluoride membranes. The following primary antibodies were used: anti-thrombin (clone 12, BD Biosciences, Franklin Lakes, NJ), anti-OPN (R&D Systems, Minneapolis, MN), anti-FAK, Phospho-FAK (Tyr397), or anti-GAPDH (Cell Signal Tech, Danvers, MA).

Statistical Analyses.

Values are expressed as the mean ± standard deviation. P < 0.05 was considered statistically significant.


Thrombin Expression Level and Its Association with the Metastatic Potential of HCC Cell Lines.

Both the mRNA and protein expression levels of thrombin in six established HCC cell lines were found to be significantly increased in comparison to nontransformed hepatic cell lines (L-02 and Chang liver cell lines) detected by real-time PCR (Fig. 1A) and western blot (Fig. 1B). Among these HCC cell lines, HepG2, PLC, and MHCC97-L cells had relatively lower invasive and metastatic capabilities (low-metastatic group), whereas MHCC97-H, HCCLM3, and HCCLM6 had higher invasive and metastatic potential (high-metastatic group). Both the mRNA and protein expression levels of thrombin in these three high-metastatic HCC cell lines were much higher than those of the three low-metastatic HCC cell lines (P < 0.01, Fig. 1A,B). These data indicate that expression of thrombin is up-regulated in HCC cell lines, and its increased expression is positively correlated with the malignant phenotype of HCC cells.

Figure 1.

Thrombin is overexpressed in highly metastatic HCC cell lines compared with HCC cell lines with lower metastatic ability or nontransformed hepatic cell lines. The six established HCC cell lines were separated into the low-metastatic (HepG2, PLC, MHCC97-L) and high-metastatic potentials (MHCC97-H, MHCCLM3, and MHCCLM6) depending on their invasive and metastatic abilities. (A) mRNA expression levels detected by real-time PCR. (B) Protein levels of thrombin detected by western blot. Both thrombin mRNA and protein levels were significantly higher in HCC cell lines than in the nontransformed hepatic cell lines (L-02 and Chang liver); the high-metastatic HCC cells had much higher thrombin levels compared with those of low-metastatic ones. *P < 0.05; **P < 0.01.

mRNA Expression Level of Thrombin in HCC Tissues and Its Association with Postoperative Tumor Relapse.

The mRNA expression levels of thrombin were evaluated in 72 human HCC tumors, in their adjacent nontumor liver tissues, and 20 normal liver tissues. The amount of thrombin mRNA was significantly increased in HCC tissues (2-&Dgr;Ct, 46.4 ± 23.5) compared with their adjacent nontumor liver tissues (19.6 ± 8.1, P = 0.005; Fig. 2A, Supporting Information Table S2). Moreover, the higher amount of thrombin mRNA in HCC tissue was significantly positively correlated with tumor recurrence (P = 0.037; Fig. 2A). These findings were confirmed by elevated thrombin protein levels in eight out of the above 72 HCC tissue samples using western blot analyses (Fig. 2B).

Figure 2.

Thrombin is expressed in HCC tissues. (A) Relative thrombin mRNA expression in nontumor and tumor tissue (left), nonrecurrent and recurrent HCC tissues (right) (mean ± SD; data from three independent experiments). (B) thrombin protein levels in eight HCC tissues (four nonrecurrent and four recurrent cases) with corresponding nontumor tissues detected by western blot. (C) Relative thrombin mRNA expression of nonrecurrent and recurrent HCC tissues in OPN+ (higher OPN expression) and OPN− (lower OPN expression) groups divided with the cutoff being the median relative OPN mRNA level.

The mRNA level of OPN in HCC tissues (2-&Dgr;Ct, 137.4 ± 69.1) were also significantly higher than that in nontumor liver tissues (2-ΔCt, 22.2 ± 7.6, P = 0.043) and normal liver tissues (19.9 ± 8.7) (Supporting Information Table S2). A scatterplot of thrombin and OPN expression revealed no correlation between thrombin and OPN mRNA levels in HCC tissues (r = −0.17, P = 0.15) (Supporting Information Fig. S1). To determine whether thrombin influences OPN-related prognosis, we divided the HCC patients into low-OPN and high-OPN groups using the median OPN expression level as the cutoff. As shown in Fig. 2C, among the HCC tissues in the high-OPN group the expression level of thrombin was significantly increased in HCCs from the patients with tumor recurrence compared with those without recurrence (P = 0.027). Importantly, this difference was not statistically significant in the low-OPN group (P = 0.457). These data indicate that thrombin might operate in conjunction with OPN in influencing the prognosis of HCC patients.

Thrombin Expression Level and Its Association with Clinicopathological Features and Prognosis of HCC Patients with Different OPN Backgrounds.

To further confirm the previous RT-PCR and western blot findings, we used immunohistochemical staining to assess the correlation between the expression levels of thrombin and OPN in HCC tumor tissues from 230 patients. We also analyzed the association of thrombin and OPN levels with HCC prognosis in the same 230 HCC patients. Positive staining for thrombin and OPN was found in 33% (77/230) and 39% (90/230) of patients, respectively. HCC tissue from 36 (15.7%) patients was positive for both thrombin and OPN (Fig. 3A).

Figure 3.

The association of thrombin and osteopontin expression detected by immunohistochemical staining with prognosis of HCC patients. (A) Different expression status of thrombin and osteopontin detected by immunohistochemical staining in the consecutive sections of HCC tissues from the same patient. HCC of case 19 is positive for both positive cases (osteopontin+/thrombin+); Case 7 is positive for osteopontin but negative for thrombin (osteopontin+/thrombin−); Case 157 is osteopontin−/thrombin+; and Case 93 is osteopontin−/thrombin−. Positive cells were stained brown (×200). Scale bar = 100 μm. (B) In Kaplan-Meier analyses, the association of thrombin expression with overall survival (OS) (left) and time to recurrence (TTR) (right) was demonstrated in whole (top), osteopontin-positive (middle), and osteopontin-negative (bottom) subgroups of HCC patients.

As shown in Table 1, thrombin-positive expression in tumor tissue was significantly correlated with tumor size (P = 0.0438), vascular invasion (P = 0.0317), and TNM stage (P = 0.0352) of HCC. However, no statistically significant association was found between the thrombin expression and other clinical characteristics. In the patients with positive OPN (OPN+), positive thrombin staining in the tumor tissue was significantly correlated with preoperative serum alpha-fetoprotein (AFP) (P = 0.0304), tumor size (P = 0.0024), vascular invasion (P = 0.0018), TNM stage (P = 0.0080), tumor differentiation (P = 0.0373), and tumor encapsulation (P = 0.0477). However, no statistically significant correlation was found between thrombin expression and these characteristics in the patients with undetectable OPN expression (OPN−) (Table 2).

The 1-, 3-, and 5-year tumor recurrence rates of those thrombin-positive (thrombin+) patients were 41.6, 67.5, and 68.8%, respectively; these tumor recurrence rates were much higher than those of thrombin-negative (thrombin−) patients (24.8, 43.1, and 47.1%, respectively; P = 0.0001). The 1-, 3-, and 5-year OS rates of thrombin+ patients (75.3, 42.9, and 40.2%, respectively) were significantly lower than those of thrombin− patients (85.6, 59.5, and 57.5%, respectively; P = 0.005) (Fig. 3B).

To further evaluate the prognostic value of thrombin for HCC patients, univariate and multivariate analyses were performed with the clinicopathological characteristics and expression of thrombin and OPN (Supporting Information Tables S3 and S4). In the univariate analysis, tumor size, vascular invasion, TNM stage, and tumor differentiation were revealed to associate with OS and TTR of HCC patients. Thrombin expression was also significantly associated with both OS and TTR and, particularly, this association was much stronger in OPN+ patients (OS, P = 0.001; TTR, P < 0.0001) compared with OPN− patients (OS, P = 0.596; TTR, P = 0.728). No significant prognostic significance was found in the other characteristics including sex, age, and hepatitis B surface antigen (HBsAg) positivity of patients for OS or TTR (Supporting Information Table S3).

Individual features that showed significance by univariate analysis were adopted as covariates in a multivariate Cox proportional hazards model and then combined variables were further analyzed. Thrombin was still revealed to be an independent prognostic indicator for both OS (P = 0.023) and TTR (P = 0.014) (Supporting Information Table S4).

We then stratified the 230 HCC patients by OPN expression and found that, in OPN+ patients, positive thrombin expression (thrombin+/OPN+) was associated with a much shorter TTR compared with those of thrombin− HCC (P < 0.0001; Fig. 3B). The 1-, 3-, and 5-year recurrence rates of thrombin+/OPN+ patients were 47.2, 86.1, and 88.9%, respectively, which were significantly higher than those of patients with thrombin−/OPN+ (20.4, 42.6, and 46.4%, respectively; P < 0.0001). The 1-, 3-, and 5-year OS rates of thrombin+/OPN+ patients (72.2, 27.8, and 22.2%, respectively) were significantly lower compared with those of thrombin−/OPN+ patients (85.2, 61.1, and 55.2%, respectively; P = 0.001). However, no significant difference in the survival and recurrence rates was observed between the thrombin− and thrombin+ patients within the OPN− group (TTR, P = 0.728; OS, P = 0.596; Fig. 3B).

Thrombin Promotes the In Vitro Proliferation and Adhesion of OPN+ HCC Cells, but Not OPN− Cells.

PLC cells were stably transfected with either wildtype OPN (PLC-OPN) or an empty vector control (PLC-CON); the OPN protein level of transfected cells was detected by western blot. PLC-OPN cells were confirmed to have an elevated level of OPN protein compared with PLC-CON cells (Supporting Information Fig. S3).

In vitro cell proliferation assays were performed to evaluate the difference between PLC-CON and PLC-OPN cells in response to thrombin treatment (2 U/mL). PLC-CON and PLC-OPN cells had similar growth kinetics in normal culture. When grown in the presence of thrombin, PLC-OPN cells demonstrated a significant increase in cell proliferation during the exponential growth phase (P < 0.05); however, PLC-CON cells demonstrated no significant proliferation changes when treated with thrombin (Fig. 4A). PLC-CON and PLC-OPN cells were evaluated for altered cell adhesion to various ECM molecules in response to thrombin (2 U/mL). Thrombin treatment significantly increased the adhesion of PLC-OPN cells, but not PLC-CON cells (Fig. 4B).

Figure 4.

Thrombin increases HCC cell proliferation and adhesion in vitro in an OPN-dependent manner. (A) Cell growth kinetics in normal culture of untreated versus thrombin-treated (2 U/mL) PLC-CON cells (left panel), and untreated versus thrombin-treated (2 U/mL) PLC-OPN cells (right panel) (n = 3 plates/timepoint). (B) Cell adhesion of PLC-CON and PLC-OPN to BSA (negative control) or adhesion strips coated with human fibronectin, vitronectin, laminin, collagen I, and collagen IV in the absence or presence of thrombin (2 U/mL), with a pretreatment period of 30 minutes. *P < 0.05; **P < 0.01; #P > 0.05.

Possible Mechanisms of Thrombin Enhancing HCC Invasion.

As shown in Fig. 5A, both recombinant N-terminal fragment and full-length human OPN were able to significantly increase the PLC-CON cell proliferation in the exponential growth phase (P < 0.05). The N-terminal fragment had a much stronger positive influence on proliferation than full-length OPN (P < 0.05). However, the recombinant C-terminal OPN fragment had no effect on PLC-CON cell proliferation (P > 0.05). To determine the effect of OPN on cell adhesion, purified OPN or its fragments were immobilized on microtiter plates and adhesion of PLC-CON cells to each recombinant protein was compared. As shown in Fig. 5B, more HCC cells adhered to N-terminal fragment of OPN compared with intact OPN (P < 0.05), whereas there was no significant adhesion to the C-terminal fragment or the negative control bovine serum albumin (BSA).

Figure 5.

Thrombin promotes proliferation of OPN-positive HCC cells through integrin β1-FAK signaling pathway. (A) Cell growth kinetics of PLC-CON cells in normal culture treated with 10 μM recombinant human OPN (n = 3 plates/timepoint). (B) PLC-CON cells (30,000/well) were allowed to attach for 1 hour to wells coated with the indicated concentration of full-length recombinant human OPN (OPN-FL), recombinant fragments (OPN-N and OPN-C), or BSA. Nonspecific cell adhesion as measured on BSA-coated wells was subtracted. (C) For PLC-OPN, the expression level of integrin-β1 mRNA was upregulated after thrombin treatment for 60 minutes. (D) PLC-OPN and PLC-CON were treated with various concentrations of thrombin for 2 hours. Protein content of cell extracts was analyzed by western blot (left panel) using anti-FAK-Tyr (P) 397 antibodies or anti-FAK antibody. These data are representative of three independent experiments. (E) FAK-Tyr (P) 397 phosphorylation induced by thrombin treatment was inhibited by 10 μg/mL of integrin-β1 neutralizing antibody AIIB2. *P < 0.05; **P < 0.01; #P > 0.05.

Proteolysis modification of OPN by thrombin cleavage has been reported to reveal cryptic binding sites for β1-containing integrins. Focal adhesion kinase (FAK) plays a critical role in integrin-β1-dependent signaling. Consistent with these previous studies, we also found up-regulated integrin-β1 mRNA expression in OPN-overexpression transfectants after thrombin treatment (Fig. 5C). We then analyzed the amount of total and phospho-FAK (Y397) in the PLC-OPN and PLC-CON cells using western blot to investigate whether OPN modification by thrombin cleavage could induce FAK activation in OPN+ HCC cells. As shown in Fig. 5D, thrombin treatment induced FAK phosphorylation in OPN-overexpression transfectants (PLC-OPN) in a dose-dependent manner. FAK was maximally phosphorylated at 2 U/mL thrombin. However, no significant FAK phosphorylation was observed after thrombin treatment of control PLC-CON cells. Our results also show that thrombin treatment did not significantly change the total protein level of FAK. Moreover, integrin-β1 neutralizing antibody AIIB2 (10 μg/mL) significantly inhibited the thrombin-induced FAK phosphorylation (Fig. 5E). These data indicate that thrombin promotes the growth and invasion of OPN+ HCC cells through the activation of the integrin-β1/FAK pathway.


Many factors, such as a patient's general condition, including liver function, macroscopic tumor morphology, and histopathological features (satellites, vascular invasion, etc.), as well as tumor stages, have proven useful in predicting the tumor recurrence and prognosis of HCC patients, and triaging the patients who need and may benefit from adjuvant therapy.22 However, these features cannot always provide exact enough information for the prediction of patient outcomes. Sometimes the patients, even though they have the same stages of disease, histopathological features of the tumor, and treatment strategy, have different clinical outcomes. Particularly, it is even harder to determine which individuals will have tumor relapse after surgical treatment in patients with early-stage HCC who do not have significant vascular invasion, regional or distant metastasis. Identification of molecular characteristics of HCC could provide supplemental information that could be useful for dividing the patients into different subgroups. This would facilitate the prediction of tumor recurrence and patient outcomes after operation, and better selection of therapeutical strategies. In this study, based on the staining of OPN and thrombin, we not only divided the HCC patients into subgroups with different prognoses, but also identified the subgroup of patients who will possibly benefit from thrombin treatment to inhibit metastasis.

The abundance of clinical and experimental evidence regarding the link between OPN and HCC metastasis makes OPN an attractive potential therapeutic target for combating HCC metastasis.1 However, direct targeting of OPN is difficult, as OPN-specific inhibitory compounds are not yet available. Another strategy for therapeutic intervention may be to target OPN indirectly by way of the signaling cascade components leading to OPN-dependent HCC metastasis. Our study is the first to identify the important role of thrombin in OPN-dependent HCC metastasis.

Previous studies have shown that thrombin-induced modification can lead to changes in OPN activity.7-10, 12, 14 Thus, investigating the role of thrombin in the OPN-mediated pathway is helpful in understanding the mechanisms by which OPN regulates the proliferation and metastasis of HCC, and in the development of a potential therapeutic target to block OPN function and control HCC metastasis.6

Previous studies have also indicated that thrombin expression is associated with a more malignant cancer phenotype.15, 16 In this study we demonstrated for the first time that the thrombin expression was significantly correlated with metastatic potential of HCC, postoperative tumor recurrence, and poor prognosis of HCC patients. This finding is supported by the fact that a high thrombin expression was significantly associated with the aggressive histopathological characteristics of HCC, such as large tumor size, vascular invasion, and high TNM staging. The prognostic value of thrombin was further confirmed to be independent of the other clinicopathological characteristics of HCC in multivariate analysis. This indicates that thrombin may serve as an independent predictor for tumor recurrence and prognosis of HCC patients.

Interestingly, the correlations of thrombin to the HCC prognosis were different in patients with various OPN expression levels. Thrombin expression was closely associated with tumor recurrence and survival in HCC patients with higher OPN levels; however, this association was not significant in those patients with lower OPN expression. The HCC patients with thrombin+/OPN+ have the poorest prognosis. These findings provide more evidence to support the fact that thrombin makes a substantial contribution, together with OPN, to HCC malignancy.

To further explore the role of thrombin in the OPN-mediated HCC metastasis, we used exogenous thrombin to treat HCC cell lines in vitro, and found that thrombin could only promote proliferation and adhesion of HCC cells with OPN overexpression (PLC-OPN). We also compared the effects of N-terminal and C-terminal fragments with intact OPN on cell growth and adhesion, and found that the OPN N-terminal fragment increased the rate of proliferation and adhesion of HCC cells to an even higher degree than intact OPN. These results suggest that OPN is necessary for the effects of thrombin on the proliferation and adhesion of HCC cells; in the other words, thrombin affects HCC malignancy through the functional roles of OPN pathway.

Previous reports have shown that proteolytic modification of OPN by thrombin cleavage not only enhances the accessibility of the binding motif (RGD) for αvβ3 integrins,11 but also reveals the cryptic binding site SVVYGLR in the N-terminal of β1-containing integrins.8, 23 In this study we found that the expression level of integrin-β1 mRNA was up-regulated in OPN-overexpressing transfectants after thrombin treatment. This suggests that thrombin collaborates with OPN to induce the increased integrin-β1 expression.24

FAK plays critical roles in β1 integrin-dependent signaling,25 in survival signaling of circulating cells to avoid anoikis,26 and to form metastatic colonies.27 Disseminated cancer cells depend on survival signals to avoid rapid elimination by apoptosis. Increasing evidence suggests that this pathway is abnormally regulated in HCC.28 To further elucidate the mechanism of these observations, we investigated the total and phosphorylated FAK levels. Treatment with thrombin significantly increased the phosphorylation of FAK in the OPN+ HCC cells; however, levels of total FAK remained unchanged. Moreover, thrombin-induced FAK phosphorylation was significantly inhibited by integrin-β1 neutralizing antibody. These data indicate that thrombin is able to regulate the integrin-β1/FAK pathway through the proteolytic modification of OPN and affect the proliferation and adhesion abilities of HCC cells.

In this study we not only provide convincing evidence that thrombin plays a crucial role in OPN-mediated function, but also an explanation as to why intravascular coagulation with generation of thrombosis has been observed in most patients with solid tumors.29, 30 A blood disorder involving hyperactivation of the coagulation system and formation of intravenous fibrin clots (thrombosis) can be the first manifestation of various tumors, including HCC.31 Meanwhile, some molecular targeted therapies such as sorafenib and sunitinib are associated with a significant increase in the risk of arterial thromboembolic events.32 The search for cancer-associated molecules responsible for thrombosis could reveal targets to fight both the side effect as well as the primary disease. The treatment should start immediately after diagnosis and in conjunction with molecular targeted therapies, especially sorafenib for those patients with advanced HCC.33 There are several thrombin inhibitors that are currently clinically available, including the broad-spectrum anticoagulants and the thrombin-specific inhibitors. Some of these agents have been demonstrated to have an inhibitory effect on metastatic behavior in experimental studies34; however, the main clinical applications of these agents thus far are for the treatment of disorders and complications, rather than for control of tumor metastasis.35 Despite these desired results, a number of unique challenges still exist for the treatment of cancer patients with antithrombotic agents, including suboptimal efficacy and high risk of bleeding using broad-spectrum agents, particularly for HCC patients, who often have a chronic hepatitis background.36 The use of more specific anticoagulants such as Argatroban, therefore, holds promise in terms of improved safety and efficacy. Meanwhile, the reversible nature and short half-life of such specific anticoagulants is considered clinically advantageous for managing patients who are at high risk of bleeding.

Based on the findings of this study, we conclude that thrombin plays an important role in OPN-mediated HCC metastasis and proliferation of HCC cells in vitro. The mechanism by which thrombin acts may be through the activation of integrin β1-FAK signaling; expression of thrombin may be helpful in the prediction of HCC prognosis; and thrombin may be a potential therapeutic target for HCC patients with tumors that overexpress OPN.