Overexpression of brachyury contributes to tumor metastasis by inducing epithelial-mesenchymal transition in hepatocellular carcinoma

Cell culture

Human simian virus 40-transformed immortal liver epithelial cells (QZG) and HCC cell lines SMMC7721, HepG2, FHCC98, and Hep3B were purchased (Bei Na Biotechnology Research Institute, Beijing, China) and preserved [23]. Cells were maintained at 37°C with 5% carbon dioxide in Dulbecco’s Modified Eagle’s Medium (DMEM) (Gibco® Life Technologies, Grand Island, USA), supplemented with 50 units/ml penicillin, 50 mg/ml streptomycin, and 10% fetal calf serum (Gibco® Life Technologies, Burlington, Canada).

Patients and tissue samples

Patients (n = 40; aged between 32 and 79 years with a mean age of 50.12 ± 15.02 years), who underwent hepatectomy for HCC between January 2006 and August 2008 in the Department of General Surgery, Xijing Hospital (Xi’an, China), were included in the Brachyury and vimentin correlation analysis. Tissue chips for HCC primary tumors were purchased from National Engineering Center for Biochip Design, Shanghai Biochip Co. Ltd., Shanghai, China. The tissue chips (n = 72) were used in the clinicopathological correlation and survival analysis. None of these patients received preoperative chemotherapy or radiotherapy. All tumor samples were derived from dissected tumor tissues and were composed of more than 90% of tumor cells without necrosis. The surgical specimens (both tumor and adjacent nontumorous tissues) were processed immediately after the operation and snap-frozen in liquid nitrogen for protein, DNA, and ribonucleic acid (RNA) extraction. All patients with HCC gave written informed consent on the use of clinical specimens for medical research. Studies using human tissues were reviewed and approved by the Committees for Ethical Review of research involving human subjects of Fourth Military Medical University, Xi’an, China.

Immunohistochemistry (IHC) and Immunofluorescence (IF)

For IF analysis, SMMC7721 and HepG2 cells were cultured on sterile glass coverslips in 24-well plates. The slides were fixed with 95% alcohol for 15 minutes at room temperature. The coverslips were washed with phosphate-buffered saline (PBS) and permeabilized for 5 minutes with 0.5% Triton X-100 in PBS. The cells were then incubated with primary antibodies including anti-E-cadherin and anti-vimentin (1:100 ratios; Santa Cruz Biotechnology, Santa Cruz, USA) after blocking with 10% normal goat or rabbit serum for 1 h. The slides were incubated with fluorescein isothiocyanate-conjugated goat anti-mouse or anti-rabbit Immunoglobulin G (IgG) as the secondary antibody at room temperature for 1 h. Slides were examined with a Nikon (Melville, USA) Eclipse TE300 fluorescence microscope, and the pictures were taken with a SPOT Diagnostic CCD camera (SPOT™ Imaging Solutions, Sterling Heights, USA).

The paraffin-embedded tissue blocks were sectioned for IHC analysis. In brief, tissue sections were deparaffinized and rehydrated. The endogenous peroxidase activity was blocked with 3% hydrogen peroxide for 10 minutes. For antigen retrieval, slides were immersed in 10 mM citrate buffer (pH 6.0) and boiled for 15 minutes in a microwave oven. Nonspecific binding was blocked by 5% normal goat serum for 10 minutes. The slides were incubated with anti-Brachyury polyclonal antibody (diluted at 1:500 ratio; Abcam, USA), anti-E-cadherin antibody, and anti-Snail polyclonal antibody (diluted at 1:200 ratio; Santa Cruz Biotechnology, Santa Cruz, USA) at 4°C overnight in a moist chamber. The slides were sequentially incubated with biotinylated goat anti-mouse IgG (1:100 dilution; Santa Cruz Biotechnology, Santa Cruz, USA) and then with streptavidin-peroxidase conjugate, each for 30 minutes at room temperature. Isotope-matched human IgG was used in each case as a negative control. Finally, the 3, 5-diaminobenzidine Substrate Kit (Dako, Glostrup, Denmark) was used for color development followed by Mayer’s hematoxylin counter staining. The slides were counterstained with hematoxylin, dehydrated and mounted. Tumor cell proportions were scored as follows: 0 (no positive tumor cells); 1 (<25% positive tumor cells); 2 (25–50% positive tumor cells); and 3 (>50% positive tumor cells). Staining intensity was graded according to the following standard: 0 (no staining); 1 (weak staining = light yellow); 2 (moderate staining = yellow brown) and 3 (strong staining = brown). The staining index (SI) was calculated as the product of the staining intensity score and the proportion of positive tumor cells. Using this method, we evaluated nuclear Brachyury expression in HCC tumor tissues and adjacent nontumorous ones by determining the SI, with scores of 0, 1, 2, 3, 4, 6, 8, 9. Only 2+ and 3+ was considered as a positive IHC result [20].

Quantitative real-time polymerase chain reaction (qRT-PCR)

Total RNA was isolated using Trizol reagent (Life Technologies, China), reverse transcribed with TaKaRa reverse transcription Kit (TaKaRa, Dalian, China) starting with 2 μg total RNA from each sample per manufacturer’s instructions. The messenger ribonucleic acid (mRNA) expressions of Brachyury and housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were analyzed by quantitative RT-PCR using a real-time LightCycler® rapid thermal cycler (Roche Molecular Biochemicals LightCycler System, Shanghai, China). Specific Brachyury primers 5’-ACTGAGAATCAGCCGGACTT-3’ (forward) 5’-CTGCACTGCAAAGAACCACT-3’ (reverse) and internal control GAPDH primers 5’-GCACCGTCAAGGCTGAGAAC-3’ (forward) and 5’-TGGTGAAGACGCCAGTGGAT-3’ (reverse) were used for mRNA amplifications. For quantitative RT-PCR, 1 μl of gene primers with SYBR® Green (Applied Biosystems, China) in 25 μl of reaction volume was used. The MTBP mRNA levels were normalized to GAPDH mRNA according to the following formula: 2^{-(CT target – CT GAPDH)}, where CT was the threshold cycle.

Plasmids and cell transfection

Human Brachyury was amplified by the PCR using cDNA from SMMC7721, confirmed by sequencing and subcloned into pCDNA3.1 expression vector(Invitrogen, Carlsbad, USA). Cell transfection was carried out by plating 2 × 10⁵ HepG2 and SMMC7721 cells on six-well plate and transfected with 800-ng constructs using Lipofectamine™ 2000 (Invitrogen Life Technology, Carlsbad, USA) per manufacturer’s instructions. Briefly, cells grew 70% to 90% confluence before transfection. The pcDNA3.1-Brachyury and pcDNA3.1-control were transfected using Lipofectamine™ 2000 in DMEM. The cells were selected for more than 4 weeks by incubation with G418 (Invitrogen, 400 ng/ml for SMMC-7721 and 600 ng/ml for HepG2) for overexpression clones. Stable single clones were selected and Brachyury expression assessed using western blotting. Transient transfection of pcDNA3.1-Brachyury and the control constructs into HCC cells were performed using Lipofectamine 2000.

Protein preparation and Western blot analysis

Harvested cells (2 × 10⁶) were placed in 1.5-ml Eppendorf tubes and homogenized with 400 μl lysis buffer (50 mM Tris-hydrochloride, pH 8.0, 150 mM sodium chloride, 0.1% sodium dodecyl sulfate, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.02% sodium azide, 100 μg/ml phenylmethanesulfonyl fluoride, and 1 μg/ml aprotinin). Cell lysates were centrifuged at 4°C for 5 minutes at 10,000 rpm, and the protein-containing supernatant was placed in fresh tubes and quantified using the Bradford protein assay.

Western blot analysis was performed as described previously [24]. In brief, the membranes with total protein were incubated with primary antibodies: anti-Brachyury and anti-Snail (1:1000 dilutions; Cell Signaling Technology, Danvers, USA), anti-E-cadherin, anti-γ-catenin, anti-vimentin, anti-Akt, anti-p-Akt, and anti-PTEN (1:200 dilution; Santa Cruz Biotechnology, Santa Cruz, USA). After repeated washing, the membranes were incubated with horseradish-peroxidase-conjugated anti-rabbit or anti-mouse secondary antibody (1:2000 dilution; Santa Cruz Biotechnology, Santa Cruz, USA). The bands were visualized using the enhanced chemiluminescence detection system (Amersham Pharmacia Biotech, Piscataway, NJ, USA) and exposed to Kodak X-OMAT film (Rochester, USA). β-actin was used as an internal sample loading control. Autoradiograms were quantified using densitometry (Bio Image Intelligent Quantifier software, Bio Image Systems Inc., Jackson, MI, USA). Relative protein levels were calculated by normalization to the amount of β-actin protein.

Wound healing and invasion assays

Wound-healing assay and invasion assay were performed as described previously [12]. Cell migration was assessed by measuring the movement of cells into a scraped; a cellular area was created using a 200-μl pipette tube, and the spread of wound closure was observed after 24 hours and photographed under a microscope. The % wound healing was quantified by the space of migration tumor cells at 24 h after scrambled /the space of wound at 0 h × 100%. Cell invasion was quantified in vitro using Transwell chambers with polycarbonate membrane filters (8-μm pore size) coated with a Matrigel™ (Sigma, St. Louis, USA). Briefly, 48 hours after transfection, cells were washed twice using DMEM and seeded in triplicate in the inner chamber of the insert containing 200 μl of serum-free medium. About 700 μl of medium containing 10% fetal bovine serum was added to the lower chamber. The plates were incubated for 24 hours at 37°C. Then, the noninvading cells from the interior of the inserts were gently removed using a cotton-tipped swab. The cells that had invaded into the bottom surface of the filter were fixed with methanol and stained with hematoxylin. The invasive ability was determined by counting the penetrating cells under a microscope at × 200 magnification on 10 random fields in each well. The experiments were performed in triplicate.

In vivo metastasis assay

Six-week-old male BALB/c nude mice (n = 10) were randomized into two groups. HepG2/Brachyury or control cells (2 × 10⁶) were injected into the tail vein of nude mice. Four mice were sacrificed 6 weeks post inoculation and consecutive sections of the whole lung were subjected to hematoxylin and eosin staining. All of the metastatic foci in lung were calculated microscopically to evaluate the development of pulmonary metastasis. The remaining mice were monitored for survival analysis.

Statistical analysis

Unless otherwise indicated, data were presented as mean ± standard deviation of three independent experiments. SPSS statistical package for Windows (Version 12.0, SPSS Inc. Chicago, IL, USA) was used for data analysis. Based on staining intensities, the Brachyury immunoreactivity was scored as negative (0 to 1) and positive (2 to 3) according to a previously reported semiquantitative scoring method [25]. The clinicopathological features of Brachyury-positive and Brachyury-negative patients were compared using Pearson χ² test for categorical variables and independent Student’s t-test for continuous data. Comparisons of three or more variables were performed using one-way analysis of variance followed by Fisher’s protected least significant difference test. Kaplan-Meier plots and log-rank tests were used for survival analysis. The DFS times were calculated from data of curative surgery to HCC recurrence, death, or the last follow-up data. For tissue microarray analysis, based on IHC scores, Brachyury protein levels in primary HCC tissues and their matched metastatic tissues were compared using Wilcoxon signed-rank test. The correlation between mRNA levels of Brachyury and Snail was analyzed using Pearson χ² test. A P value less than 0.05 was considered statistically significant.

Expression of brachyury is upregulated in human HCCs

To explore the role of Brachyury in HCC development, the expression of Brachyury in various human HCC cell lines were evaluated. As shown in Figure 1A, elevated expression of Brachyury was observed in all four HCC cell lines compared with that in QZG cell lines. Brachyury mRNA was significantly increased in HCCs relative to paired noncancerous tissues of 40 patients (Figure 1B), which was further confirmed by Western blot assay (Figure 1C,D). Immunohistochemical analysis showed little positive staining of Brachyury in nontumorous liver cells, but strong nuclei staining of Brachyury (upregulated) in 62.5% (70/112) of the patients with HCC. Brachyury was predominantly located in the nuclei of hepatocytes and tumor cells (Figure 1E). Brachyury positive expression was found in tumor tissues at a value (62.5%, 70 of 112 patients) higher than that in adjacent cirrhosis tissues (23.2%, 26 of 112 patients) (P < 0.05).

Brachyury overexpression predicts poor prognosis of HCC

To investigate the clinical significance of Brachyury overexpression in HCC, microarray analysis of HCC tissues from 72 patients who underwent liver resection and 40 PPEP tissues was performed (Table 1). The Mann-Whitney U test was used to assess the correlations between the staining intensity of Brachyury protein and clinicopathologic variables of HCC. As shown in Table 2, the Brachyury expression in patients with HCC was correlated with tumor size (P = 0.037), intrahepatic invasion (P = 0.025), distance metastasis (P = 0.039). However, the Brachyury protein staining intensity showed no significant relationship with gender, age, α-fetoprotein and HBV infection. Based on the results from IHC, all 112 patients with HCC were divided into two groups: Brachyury-positive expression group (n = 70) and Brachyury-negative expression group (n = 42). Patients in positive expression group exhibited shorter 5-year overall survival (OS, median OS were 16 and 43 months, respectively, difference of 27 months, P = 0.004) than those in negative expression group (Figure 2A). Consistently, the 3-year OS rate after surgery was much lower in Brachyury-positive group than that in Brachyury-negative group (Figure 2B). Thus, Brachyury overexpression could serve as a valuable predicting factor for recurrence and poor survival of patients with HCC.

Characteristic	No. of patients
Patients	112
Gender
Male	92
Female	20
Age
< 55	72
≥ 55	40
AFP (μg/L)
< 100	47
≥ 100	65
HBV
Positive	91
Negative	21
Liver Cirrhosis
Yes	66
No	46
Tumor size (cm)
< 3	37
≥ 3	75
Intrahepatic metastasis
Y	41
N	71
Distant metastasis
Y	27
N	85
TNM stage
I	27
II	62
III	23

Features		Relative Brachyury expression		p Value
		High	Low
Sex	Male	60	32	0.797
	Female	12	8
Age	<55	45	27	0.683
	≥55	27	13
AFP	<100	32	15	0.551
	≥100	40	25
HBV	Positive	58	32	0.807
	Negative	13	8
Liver Cirrhosis	Yes	47	19	0.075
	No	25	21
Tumor size (cm)	<3	19	18	0.037
	≥3	53	22
Intrahepatic Metastasis	Y	32	9	0.025
	N	40	31
Distant Metastasis	Y	22	5	0.039
	N	50	35
TNM Stage	I	15	12	0.401
	II	40	22
	III	17	6

Brachyury enhances invasive and metastatic potential of HCC

The EMT promotes tumor progression by increasing the invasiveness and migratory capacity of the tumor cells. To test whether Brachyury-SMMC7721 and Brachyury-HepG2 cells acquired greater migratory and invasive capabilities, in vitro scratch wound healing and Matrigel invasion assays were performed. The Matrigel invasion assay showed that the invasiveness of the Brachyury-expressing cells was significantly higher than empty vector-transfected cells (P < 0.001, independent Student’s t test; Figure 3A). Similarly, the wound healing assay demonstrated that the ectopic expression of Brachyury increased the cell motility in both Brachyury-SMMC7721 and Brachyury-HepG2 cells compared with Vec-SMMC7721 and Vec-HepG2 cells (Figure 3B). To further verify the effect of Brachyury on tumor metastasis, in vivo metastasis assay was performed in nude mice. Brachyury-HepG2 cells were injected into nude mice via tail vein, Empty vector trasfected cells were used as controls. Six weeks later, more and larger micrometastatic lesions were microscopically detected in the lungs of nude mice inoculated with Brachyury-HepG2 compared to that in the control group (P < 0.001, independent Student’s t test; Figure 3C). These data suggested that Brachyury promoted HCC invasion and metastasis.

Brachyury promotes EMT in HCC cells

As EMT has been accepted as a potential mechanism underlying cancer metastasis, we explored the role of Brachyury in regulating EMT in HCC cells, Up-regulation of Brachyury in HepG2 cells resulted in the decreased expression of epithelial markers (E-cadherin and γ-catenin) and increased expression of mesenchymal marker (snail), as evidenced by immunofluorescence staining (Figure 4A). Western blot analysis also revealed that the protein level of vimentin (mesenchymal marker) was increased, while E-cadherin and γ-catenin were decreased in Brachyury-overexpressing cells compared with control cells (Figure 4B).

Brachyury regulates EMT via PTEN/AKT/Snail pathway

Numerous studies have suggested that Akt activation plays a pivotal role in tumor progression via induction of EMT [26]-[29]. Thus, the Akt activation is usually up- and downregulated by tumor suppressor PTEN. Western blot revealed that PTEN expression was significantly influenced by Brachyury overexpression (Figure 5A), and it confirmed the involvement of PTEN in Brachyury-mediated Akt activation. Snail has been documented to be regulated by Akt and is a predominant mediator of EMT [30],[31]. To test if Snail was involved in Brachyury-mediated EMT, the effect of Brachyury on Snail activation was examined. As expected, expression of Snail was significantly increased in cells with Brachyury-overexpression.

The Akt/Snail pathway involvement in EMT-mediated Brachyury-overexpressed HCC cells was investigated. Brachyury-HepG2 cells were treated with Akt inhibitor (LY290004) and small interfering RNA (siRNA) against Snail, and the effects on cell EMT were determined as described previously. As shown in Figure 5, treatment with Akt inhibitor prevented the Brachyury-mediated increase in p-Akt and Snail (Figure 5B). Transfection with siRNA against Snail effectively abolished Brachyury-induced Snail expression (Figure 5C). The expression of E-cadherin was correspondingly increased while vimentin was reduced in Brachyury-transfected cells after treatment with Akt inhibitor or Snail siRNA. These results indicate that Brachyury mediates the mesenchymal transition of HCC cells through the Akt/Snail pathway.