Human tumor necrosis factor (TNF)-alpha-induced protein 8-like 2 suppresses hepatocellular carcinoma metastasis through inhibiting Rac1

Our previous researches have showed that human TIPE2 was downregulated in tumor tissues of HCC compared with the paired adjacent non-tumorous tissues [10], suggesting it may be associated with HCC. Here we firstly analyzed the correlation of TIPE2 expression to clinical pathological features of HCC. As shown in Table 1, there was no significant correlation of TIPE2 expression to age, gender, cirrhosis, hepatitis B, serum AFP and pathological grade (p > 0.05). However, the TIPE2 had a significant relation to TNM stage. Rate of TIPE2 loss or low expression was higher in tumors from patients with TNM stage III-IV than that with TNM stage I-II (p = 0.044). Since TNM stage III-IV means that tumor is accompanied with vascular invasion or distant metastasis, the data suggests that the loss or low expression of TIPE2 may be associated with metastasis of HCC.

To explore the role of TIPE2 in HCC, we further detected the effect of TIPE2 on growth, colony formation, migration and invasion of HCC cells in vitro. As shown in Additional file 1: Figure S1, cells untransfected (Control) or transfected with empty plasmid (Mock) showed undetectable TIPE2 expression both on mRNA level by RT-PCR and protein level by western blot while cells transfected with TIPE2 recombinant plasmid (TIPE2) revealed high level of TIPE2 mRNA and protein. Further, overexpression of TIPE2 suppressed significantly cell viability from 48 hours and colony formation compared with Mock cell in both BEL-7402 and HepG2 (Additional file 1: Figure S1B and C). Overexpression of TIPE2 reduced markedly migration and invasion of BEL-7402 and HepG2 compared with that of Mock (Figure 1). This suppressive effect of TIPE2 was also detected in HCCLM3 with high metastasis ability (Additional file 2: Figure S2).The results indicate that TIPE2 can effectively suppress malignant phenotypes of HCC cells, especially the migration and invasion capacity in vitro.

To provide in vivo evidence that TIPE2 suppresses HCC tumor growth and metastasis, we established two xenograft tumor models in nude mice, subcutaneous xenograft tumor model and liver orthotopic transplanted tumor model using HCCLM3 cells. In subcutaneous xenograft tumor model, we found that treatment with TIPE2 plasmid attenuated remarkably the subcutaneous tumor growth (Figure 2A). The overall mean tumor volume of TIPE2 group was much smaller than that of Mock group (Figure 2B-D). Consistent with the results, the mean tumor weight and concentration of serum AFP in TIPE2 group were also significantly lower than those of Mock group (Figure 2E and F). Moreover, the results from IHC and H&E staining showed that tumor in TIPE2 treatment group had obvious TIPE2 expression and revealed higher differentiation compared with those in the Mock group (Additional file 3: Figure S3). These results indicate that re-expression of TIPE2 in the HCC cells attenuates the tumor growth in vivo. Furthermore, we analyzed the effect of TIPE2 on the metastasis of HCC in the liver orthotopic implantation tumor model. As shown in Figure 3, 35 days after inoculation of xenograft tumor, tumor volume, weight and serum AFP in TIPE2 group obviously reduced compared with those in Mock group (Figure 3A-D). More importantly, 100% (5/5) of Mock group mice were detected spontaneous lung metastasis while only 60% (3/5) of TIPE2 group mice did. In addition, tumors had invaded into the pancreas, diaphragm and abdominal wall in Mock group but not in TIPE2 group (Figure 3E and Table 2). These results demonstrate that TIPE2 markedly suppresses growth and metastasis of HCC cells in vivo.

The results above suggest that TIPE2 is able to inhibit metastasis and invasion of HCC in vitro and in vivo. Next, we want to know how TIPE2 impact them. Our previous report has demonstrated that murine TIPE2 is able to inhibit F-actin polymerization and cell migration by targeting RGL, one of Ras signaling pathway effectors [10]. However, we found that expression of RGL was very low in all HCC cell lines detected (Additional file 4: Figure S4A), suggesting RGL may be not a major target of TIPE2 in HCC. More recent research showed that TIPE2 regulated negatively murine innate immunity by binding and blocking Rac1GTPases [8]. So we hypothesized that TIPE2 may suppress the migration and invasion via targeting Rac1 in HCC. To test this hypothesis, we undertook approaches as follows. First, we detected expression of Rac1 and its action in migration and invasion of HCC. Consistent with previous researches, high level of Rac1 expression were detected in multiple cell lines and its expression was positively related with metastasis capacity (Additional file 4: Figure S4A). To determine whether TIPE2 regulates migration and invasion through Rac1, we manipulated the Rac1 activity using chemical blocker NSC23766. The Rac1 antagonist NSC23766 attenuated cell migration and invasion markedly, and effectively reduced the difference between mock and TIPE2 group at 50 μM both in BEL-7402 and HepG2 (Additional file 4: Figure S4C). Consistent with this, silence of Rac1 expression by specific siRNA abolished the inhibition effect of TIPE2 on migration and invasion of tumor cells (Figure 4A). Furthermore, coimmunoprecipitation (co-IP) revealed that human wild type TIPE2 could bind to Rac1 while mutant TIPE2 lost the ability in HCC cell (Figure 4B). Additionally endogenous human TIPE2 was also able to bind to Rac1 in human monocyte cell line Thp-1 (Additional file 4: Figure S4B). Using a GST pull-down assay, we found that transfection of wild type TIPE2 decreased the Rac1GTPase activity while mutation of TIPE2 reversed its inhibition effect, indicating that human TIPE2 can also bind and block Rac1 activity in HCC (Figure 4C). Third, mutation of TIPE2 reversed inhibiting effect of wild type TIPE2 on migration and invasion both in BEL-7402 and HepG2 (Figure 4D). All the results indicate that TIPE2 suppresses migration and invasion of HCC cells via inhibiting Rac1 pathway.

It has been known that microfilaments play a key role in cytoskeleton reorganization and in cell motility and tumor metastasis potential [13]. Therefore, we examined the organization of microfilaments in cells transfected with Mock, wild type TIPE2 or mutant TIPE2 plasmid by immunofluorescence method. Compared with Mock group, there was an obvious microfilament depolymerization in TIPE2 group (Figure 5A) but this action was reversed by TIPE2 mutation. In addition, the MMP9 and uPA are responsible for the degradation of extracellular matrix components and play important roles in the process of cancer invasion and metastasis [14, 15]. Therefore, we further examined the effect of TIPE2 on expression of MMP9 and uPA expression. As shown in Figure 5B and C, wild type TIPE2 obviously attenuated MMP9 and uPA expression on mRNA and protein level but TIPE2 mutation lost this ability. Additionally, we found that expression of MMP9 and uPA also decreased after silence of Rac1 (Additional file 4: Figure S4 D, E). All the results indicate that TIPE2 decreases F-actin polymerization and expression of MMP9 and uPA via inhibiting Rac1 pathway.

One hundred and twelve primary hepatocellular carcinoma specimens were used for detection of TIPE2 expression. Of them, 77 specimens were from HCC tissue chip (OUTDO BIOTECH, Shanghai, China). Other 35 specimens were obtained from patients who underwent operations at Qilu Hospital of Shandong University. These human procedures were preapproved by the Institutional Review Board of the Shandong University and were approved by the Institutional Review Board of the Shandong University.

The human HCC cell lines, BEL-7402, HepG2, SMMC-7721 and human monocyte cell line Thp-1 were purchased from Shanghai Cell Bank of Chinese Academy of Sciences (Shanghai, China), grown in RPMI 1640 medium (Gibco, CA, USA) supplemented with 10% heat-inactivated fetal bovine serum (FBS) (Gibco). HCCLM3 cell line was purchased from Liver Cancer Institute, Zhongshan Hospital (Shanghai, China), grown in DMEM medium (Gibco) supplemented with 10% FBS.

Full-length human TIPE2 was generated from the cDNA clone by PCR and cloned in frame with a C-terminal Flag into vector PRK5.The mutant TIPE2 in which the TIPE2 N-terminal lysine or arginine residues, Lys-15, Lys-16, and Arg-24 were replaced with glutamine or alanine was generated by PCR-based site-directed mutagenesis as previously described [8]. Specific siRNA for Rac1 and nonspecific negative control were purchased from Sigma-Aldrich (Louis, USA). Transfection of tumor cells with plasmid or siRNA was performed using Lipofectamine 2000 according to the manufacturer’s protocols (Invitrogen, Carlsbad, CA, USA).

Total RNA was extracted from cells using Trizol Reagent (Invitrogen). Semi-quantative RT-PCR was performed using 2 × Taq MasterMix (CWBIO, Beijing, China). Real-time PCR was performed using UltraSYBR Mixture (CWBIO). The sequences of the sense and antisense primers were as follows: MMP9: 5′-GCATTCAGGGAGACGCCCATTT AACGACA-3′, and 5′-CTGACACTCCCGGTGGG AAATCA-3′; TIPE2: 5′-ACTGA GTAAGATGGCGGGTCG-3′, and 5′-TTCTGGCGAA AGCGGGTAG-3′; Rac1: 5′-AT GTCCGTGCAAAGTGGTATC-3′, and 5′-CTCGGATCGCTTCGTCAAACA-3′; GAPDH: 5′-AACGGATTTGGTCGTATTGGG-3′, and 5′-CCTGGA AGATGGTGAT GGGAT-3′. Relative levels of gene expression were determined with GAPDH as the control.

Equal amount of protein was separated by SDS-PAGE and transferred onto PVDF membranes (Millipore, Billerica, MA, USA). Membranes were probed overnight at 4°C with the following primary antibodies: rabbit polyclonal antibody against human TIPE2 (1:300; BOSTER, Wuhan, China), rabbit monoclonal antibody against the matrix metalloproteinase 9 (MMP-9) and urokinase-PA (u-PA) (1:1000; EPITOMICS, Hangzhou, China), mouse monoclonal antibody against RGL1 (1:200; Santa Cruz, CA, USA), Rac1 (1:300; abcam, Hongkong), β-actin (1:1000; ZSGB-Bio, Beijing, China), followed by secondary antibodies (1:2000; goat anti rabbit or mouse IgG, ZSGB-Bio) conjugated with peroxidase for 1 h at room temperature. After washing, signals were visualized by eECL Western Blot Kit (CWBIO).

The paraffin slides were stained with rabbit antibody against TIPE2 (1:200) at 4°C overnight. Secondary staining was performed with HRP-conjugated anti-rabbit IgG using a MaxVsion Kit and 3, 5-diaminobenzidine (DAB) peroxidase Substrate Kit (Maixin Co., Fuzhou, China). The sections were counterstained with hematoxylin. Isotope-matched human IgG was used as a negative control. All IHC staining was independently assessed by two experienced pathologists. The staining intensity was scored from 0 to 3 (0, no staining; 1, weak; 2, moderate; 3, strong). The staining extent was scored from 0 to 3 based on the percentage of positive cells (0, < 1%; 1, 1%-33%; 2, 34%-66%; 3, 67%-100%). The two scores for each slide were then combined to produce a final grade of TIPE2 expression: 0, total score = 0; 1+, total score = 1–2; 2+, total score = 3–4; 3+, total score = 5–6. When there were discrepancies between the two pathologists, the average score was used.

Cells were seeded in six-well plates at a density of 1500 cells per well for 1–2 weeks and then fixed with 20% methanol and stained with 1% crystal violet. Colonies that is consisted of more than 50 cells were counted and calculated as a percentage of that of the control group. The experiment was independently performed for three times.

Cells were seeded in 96-well plates at 5000 cells/well and cultured for indicated time points. Cell viability was evaluated using CCK8 (Beyotime, Haimen, China) according to manufacturer’s instructions. The absorbance was determined at 450 nm wave length. Each time point was repeated in three wells and the experiment was independently performed for three times.

Tumor cell migration and invasion were analyzed in 24-well Boyden chambers with 8-μm pore size polycarbonate membranes (Costar, Acton, USA). For invasion assay, the membranes were precoated with 50 μg Matrigel (BD Biosciences, San Diego, USA) to form matrix barriers. Cells (1 × 10⁵) were resuspended in 100 μl serum-free medium and placed in the upper chamber, and the lower compartments were filled with 600 μl medium with 10% FBS. After incubation, the cells remaining on the upper surface of the membrane were removed. The cells on the lower surface of the membrane were fixed and stained with crystal violet and counted under a light microscope at ×200 magnification. In some experiments, Rac1 inhibitor, NSC23766 (Calbiochem, San Diego, USA) was used to inhibit Rac1 activity.

The cells on the cover slip were fixed, permeabilized and then stained with Tetramethylrhodamine (TRITC)-conjugated phalloidin (Sigma–Aldrich, Louis, USA) for 1 h. Nuclei were stained by 4′,6-diamidino-2-phenylindole (DAPI) (Beyotime) for 5 min. Results were analyzed on a confocal laser microscopy (Carl Zeiss, LSM780, Oberkochen, Germany).

Male athymic BALB/c nu/nu mice (4–6 week old) were purchased from Chinese Academy of Sciences (Shanghai, China) and maintained in laminar-flow cabinets under specific pathogen-free conditions. For evaluation of the tumor growth in vivo, 1 × 10⁷ HCCLM3 cells in 100 μl of PBS were injected subcutaneously into flank of nude mice (n = 10). When tumors reached approximately 1 mm³ in size 10 day later, these mice were randomly divided into two groups, Mock group (n = 5) was treated with 20 μg of empty plasmid and TIPE2 group (n = 5) with TIPE2 plasmid by intra-tumor injection every 3 days, respectively. The tumor size was measured and the tumor volume was calculated as follows: length × width × width × 0.4. After another 25 days, the implanted tumors were dissected and cut into pieces of around 1 × 1 × 1 mm³ and transplanted into liver parcel of other nude mice (n = 5 for each group). The mice were sacrificed 35 days after innoculation. All possible metastasizing visceral organs or tissures including lungs, pancreas, diaphragm and abdominal wall were removed and processed for standard histopathological study. Serial sections were made for every tissure. Any slide with metastasis was assumed as positive. The serum AFP was detected on electro-chemiluminescence immunoassay systerm (Roche, Cobas E601, Germany).

Cell lysate was prepared as previously described [8]. One μl mouse monoclonal antibody against Flag (HuaAn, Hangzhou, China), Rac1 or isotype IgG was added to 1 ml of cell lysate and incubated for 1 h at 4°C and 20μl of resuspended Protein A/G Plus-Agrose (Santa Cruz) was added to the above mixture and incubated for 4 h at 4°C with rotation. The pellet was washed four times with PBS and boiled in 2× Laemmli buffer. The proteins were detected by western blot.

Cell lysate was prepared as previously described [8]. The lysate was incubated with 20 μg of p21-activated kinase (PAK)-GST protein beads (Cytoskeleton) for 30 min at 4°C. After washing, proteins on beads and in total cell lysates were subjected to western blot to determine the level of active Rac1.

Statistical analysis was performed with SPSS 13.0 (SPSS Inc., Chicago, USA). The wilcoxon test was employed to compare qualitative variables while the Student t test for quantitative variables. All statistical tests were two-sided and P value < 0.05 was considered statistically significant for all tests.