Thyroid hormone enhanced human hepatoma cell motility involves brain-specific serine protease 4 activation via ERK signaling

Several isogenic HepG2 cell lines stably expressing high levels of wild-type TRα1 and TRβ1 (HepG2-TRα1 and HepG2-TRβ1, respectively) were established. Specifically, four HepG2 cell lines, HepG2-TRα1#1, HepG2-TRα1#2, HepG2-TRβ1#1 and HepG2-Neo, and J7-TRα1, and parental cells expressing various levels of TR (Figure 1A) were used for analyses. Notably, BSSP4 expression was stimulated by T₃ in the HepG2-TRα1#1, HepG2-TRα1#2 and HepG2-TRβ1 cell lines at both the mRNA (Figure 1B) and protein levels (Figure 1C) in a time- and dose-dependent manner. In contrast, T₃ had a marginal or no effect on BSSP4 expression in HepG2-Neo cells (Figure 1B, C). Moreover, we have examined the BSSP4 regulation by T₃ in parental cells expressing endogenous TR, such as J7 (Figure 1D, I), Huh7 (Figure 1D, II) and SK-Hep-1 cells (Additional file1: Figure S1B). We have depleted the TRα and TRβ expression with siRNA in SK-Hep1 cell (Additional file1: Figure S1A). The BSSP4 regulation was abolished significantly after silencing the TRα and TRβ expression (Additional file1: Figure S1B). Several HCC cell lines expressing low levels of endogenous TR proteins can be induced the BSSP4 expression with a lesser extent after T₃ application. According to the data, we conclude the BSSP4 regulation by T₃ can be observed in the parental cells as well as TR overexpression system.

The reporter assay was performed to determine the position of the thyroid hormone response element (TRE) to further clarify the regulatory effects of T₃ on BSSP4 at the transcriptional level. The BSSP4 5′-flanking region encompassing nucleotides -2066/-7 (relative to the transcription initiation site) with numerous predicted putative TREs (Figure 1E) was cloned and inserted upstream of the luciferase reporter gene in pGL2-luc (Construct p1) to generate Construct p2. The pA3TK-luc construct containing a minimum thymidine kinase promoter was designated Construct p6. Serial deletion mutants were additionally generated (Figure 1E). The transcriptional activities of the BSSP4 promoter fragments are illustrated in Figure 1E. Among these, only the p10 construct containing two putative TREs was activated about 3.5-fold by T₃ in HepG2-TRα1 cells. The two TREs in the p10 fragment were sequentially mutated to yield p12 and p13 constructs. However, upon mutation of the other putative TRE (pal), luciferase activity of the p12 construct was completely abolished (Figure 1E). These findings suggest that T₃ regulates BSSP4 at the transcriptional level by binding to the putative TRE site between positions -628/-498 (p10) encompassing a pal-like sequence between positions -609 to -594 (AGGTCCTTGCTGTCCT).

To further determine whether BSSP4 TRE (pal) is directly targeted by TR proteins, the ChIP assay was performed. TR proteins were clearly associated with the TRE region of the BSSP4 promoter in vivo both HepG2-TRα1 and J7-TRα1 cells (Figure 1F). Notably, TRα1 and RXRα were recruited to the TRE-binding site (Figure 1F, lanes 3, 4), whereas control IgG produced only background levels (Figure 1F, lane 2). Accordingly, we propose that the TRα1 and RXRα complexes bind to the BSSP4 promoter for transcriptional regulation.

We detected the BSSP4, TRα and TRβ expression in six hepatoma cell lines (Additional file2: Figure S2). The positive correlation of BSSP4 vs. TRα/TRβ expression was observed in most of the hepatoma cell lines. To determine the specific functions of BSSP4, control cell lines (Figure 2A, Huh7-control #1 and #2) and those overexpressing BSSP4 (Figure 2A, Huh7-BSSP4 #1 and #2) were established. BSSP4 has been detected in several cancer types[9]. The protein was originally identified as a member of the human serine protease family, also designated tryptase ϵ[7]. In view of several earlier findings that proteases function in extracellular matrix remodeling during cancer cell progression and development[10], we aimed to ascertain whether BSSP4 also plays a role in cancer processes.

The proliferation rates of the two stably overexpressing BSSP4 cell lines (Huh7-BSSP4 #1 and #2) were similar to those of the two control cell lines (Huh7-control #1 and #2), as shown in Additional file3: Figure S3A. However, cell lines overexpressing BSSP4 displayed significantly increased (~3 to 6-fold) migration (Figure 2B) and invasion (Figure 2C), compared with control cells. Our results indicate that BSSP4 functions in cell migration and invasion, but has no effect on cell growth. Furthermore, after depleted BSSP4 in Mahlavu cells, the migration ability of Mahlavu-BSSP4-depleted cells was decreased compared with Mahlavu–control cells by the Transwell assay (Figure 2D, E). Based on the functional assay in BSSP4-depleted and BSSP4-overexpressed cells, BSSP4 has the ability to accelerate tumor cell migration.

To further clarify the mechanisms involved in BSSP4-regulated cell migration and invasion, we selected several markers of epithelial-mesenchymal transition (EMT) for examination[11]. Figure 3A presents BSSP4 expression levels in Huh7-BSSP4 and Huh7-control cells. The N-cadherin level was increased and E-cadherin decreased in BSSP4-overexpressing cells (Huh7-BSSP4 #1 and #2), compared with control cells (Huh7-control #1 and #2), on a Western blot (Figure 3B). Based on these results, we propose that the EMT process is involved in BSSP4-mediated cancer cell progression.

Previously, Yasuda et al.[6] demonstrated that BSSP4 is a serine protease that catalyzes the progression of zymogen pro-uPA to form active uPA. Moreover, the uPA-uPAR system has been linked to the mitogen-activated protein kinase (MAPK)/extracellular signal-regulated kinase (ERK) signaling pathway[12‐14]. ERK kinases are frequently abnormally activated in most human cancers, leading to proliferation and acceleration of other oncogenic processes, such as angiogenesis and survival[15]. Accordingly, we examined whether the ERK pathway is implicated in BSSP4-induced phenotypes. Our results showed marked upregulation of phosphate-ERK in BSSP4-overexpressing cells (Huh7-BSSP4 #1 and #2), compared with control cells (Huh7-control #1 and #2) (Figure 3C). A number of groups have identified CCAAT/enhancer-binding protein beta (C/EBPβ and vascular endothelial growth factor (VEGF) as ERK downstream genes[16]. VEGF facilitates the formation of new vessels in angiogenesis and plays a critical role in tumor metastasis. Its expression is highly correlated with tumor cell metastasis potency in various cancers[17]. Similarly, C/EBP plays diverse roles in multiple cellular processes, such as the cell cycle, extracellular signaling, tissue development, and is implicated in cancer processes[18]. Subsequently, we investigated whether C/EBPβ and VEGF are targeted by BSSP4. Notably, C/EBPβ was upregulated in the nucleus and VEGF was extensively secreted into the medium in BSSP4-expressing cells (Huh7-BSSP4 #1 and #2), compared with their control counterparts (Huh7-control #1 and #2) (Figure 3C, D).

We further examined whether this BSSP4-mediated mechanism occurs in T₃-regulated hepatoma cells. Our results showed upregulation of nuclear phosphate-ERK and C/EBPβ after T₃ stimulation in both HepG2-TRα1 and J7-TRα1 cells (Figure 3E). Additionally, VEGF secretion into the medium was increased in T₃-treated HepG2-TRα1 and J7-TRα1 cells (Figure 3F). Accordingly, we conclude that stimulation of ERK, C/EBPβ and VEGF expression by T₃ in HepG2-TRα1 or J7-TRα1 cells is mediated via BSSP4.

In view of the above results, we hypothesized that T₃ regulates BSSP4 through the ERK-C/EBPβ-VEGF cascade, leading to cancer cell progression. To examine this possibility, we treated stable Huh7-BSSP4 cell lines with DMSO and U0126, a MEK inhibitor[19, 20], followed by the migration and invasion assays. The migration and invasion ability of BSSP4-overexpressing cells (Huh7-BSSP4 #1 and #2) was increased about 4-fold, relative to control cells (Huh7-control #1) (Figure 4A, B). However, this increase was abolished by approximately 25% upon U0126 treatment of BSSP4-expressing cells (Huh7-BSSP4 #1 and #2) (Figure 4A, B). The quantified results are shown in Figure 4A, B, lower panel. Based on these findings, we propose that the ERK pathway is implicated in BSSP4-mediated cell migration. We further ascertained whether this phenomenon occurs in T₃-regulated hepatoma cells. As expected, the migration and invasion ability of J7-TRα1 cells was enhanced (~3.5-fold) after T₃ stimulation (Figure 4C, D), which was attenuated (~30%) upon U0126 treatment (Figure 4C, D). The quantified results are illustrated in Figure 4C, D, lower panel. Thus, BSSP4 appears to mediate cell migration through the ERK signaling pathway.

We observed upregulation of p-ERK, C/EBPβ and VEGF in both BSSP4-overexpressing and T₃-treated hepatoma cells. Moreover, the ERK signaling pathway is clearly involved in BSSP4- and T₃-modulated cell migration ability. This led us to investigate whether C/EBPβ and VEGF are influenced by blocking the ERK signaling pathway. Our results showed that expression levels of nuclear p-ERK and C/EBPβ are decreased in stable Huh7-BSSP4 cells after U0126 treatment, compared with those in control cells (Figure 5A, lanes 7, 8 vs 3, 4). VEGF secretion into the medium of stable Huh7-BSSP4 cells was additionally attenuated following U0126 treatment (Figure 5B, lanes 7, 8 vs 3, 4). We further determined whether BSSP4 has a similar influence in T₃-treated hepatoma cells. Interestingly, in both T₃-treated HepG2-TRα1 and J7-TRα1 cells, expression of p-ERK and C/EBPβ in the nucleus was blocked after U0126 treatment, compared with control cells (Figure 5C, D). Furthermore, VEGF secretion into the medium was significantly attenuated following U0126 treatment in both T₃-treated cell lines (Figure 5E, F). Furthermore, the p-VEGFR and VEGFR expression were determined in BSSP4- and TR-overexpressing cells. The p-VEGFR level was increased in BSSP4-overexpressing cells. Consistently, the p-VEGFR expression was up-regulated after T₃ treatment in the HepG2-TRα cell (Additional file4: Figure S4). A Similar result can be observed in J7-TRα cell (data not shown). Therefore, we speculate that BSSP4 influenced tumor motility may through VEGF-VEGFR cascade.

The relationship between BSSP4 and cancer-related molecules and pathways was investigated using Human Cancer PathwayFinder RT Profiler metastasis-associated PCR Arrays (Quiagen; Kowloon, Hong Kong). We found 20 genes were altered (about 1.5 ~ 3 fold) in BSSP4 overexpressing cells. The BSSP4-regulated molecules are divided into several categories based on the functions such as cytokines and chemokine pathway (IL-18, CCL7, CXCL12), cell to cell adhesion pathway (PNN, SYK, MCAM), metastasis-related genes (GNRH1, TIMP2, KISS1R, TSHR, TRPM1, SSTR2), transcription factors and regulators (RORB, NR4A3, SMAD2, SMAD4), ECM cleavage pathway (MMP7) and cell cycle regulation (PTEN) were altered (Additional file5: Figure S5). Thus, BSSP4 may play a role in cancer progression via alteration of related genes. Our results collectively indicate that BSSP4 is upregulated by T₃, and subsequently activates the ERK-C/EBPβ-VEGF cascade to promote cancer cell progression.

To verify whether the in vitro effect of BSSP4 also occurs in vivo, SCID mice were injected with Huh7-BSSP4 and Huh7-control cells. However, no tumors were observed in the lung and liver in either cell line. Based on these findings, we speculated that the Huh7 hepatoma cell line is non-tumorigenic. Therefore, a stable J7-BSSP4 cell line was established. Notably, the BSSP4 level was increased in two stably expressing BSSP4 clones (J7-BSSP4 #1 and #2), compared to two control cell lines (J7-control #1 and #2) (Figure 6A). Similar to Huh-7 cells, proliferation was not influenced in stable BSSP4-overexpressing J7 cell lines, compared with controls (Additional file3: Figure S3B). Moreover, BSSP4-overexpressing cells (J7-BSSP4 #1 and #2) showed higher migration and invasion abilities, relative to their control cell counterparts (J7-control #1 and #2, respectively) (Figure 6B, C). The in vitro phenotypes were similar between Huh7-BSSP4 and J7-BSSP4 cell lines (Figures 2 and6). Additionally, we have established J7-TR-BSSP4 knockdown (KD) cell lines. The BSSP4 expression in control cells was induced by T₃, however, the induction was abolished in the J7-TR-BSSP4 KD cells (Figure 6D, E). Obviously, the migration ability of J7-TR-control cells was enhanced by T₃ stimulation, however, the effect was attenuated in the J7-TR-BSSP4 KD cells (Figure 6E) restoring a more “normal” phenotype. The migration ability was significantly (p < 0.018) restored in J7-TR-BSSP4-KD-T₃ cell line after T3 treatment compared to the BSSP4 silencing J7-TR-BSSP4-KD cell line (Figure 6E, right panel). Based on the evidences, the migration phenotype in the T₃-treated hepatoma cells can be restored in a BSSP4-KD condition.

SCID mice were employed to examine whether the in vitro effects of J7-BSSP4 can be replicated in vivo. Significantly, in SCID mice, J7-BSSP4 cells formed higher numbers of lung foci (Figure 7A) and had a higher metastatic index (Figure 7B), compared with control cells, as well as elevated BSSP4 expression, as evident from H & E staining (Figure 7C, D, upper panel) and IHC (Figure 7C, D, lower panel) analyses, respectively. Based on these findings, we conclude that BSSP4 promotes cell migration and invasion in J7 hepatoma cells, both in vitro and in vivo.

Next, we examined the clinicopathologic significance of BSSP4 expression in HCC and evaluated the correlation between TRs and BSSP4. Overall, 74 consecutive HCC patients were enrolled, and their BSSP4 levels analyzed using Western blot. Among the 74 HCC sample pairs, the BSSP4 was overexpressed in 52.7% (39 of 74) cancerous tissues, compared with matched noncancerous tissues. Furthermore, both TRα1 and TRβ1 levels were elevated by about 29.7% (22 of 74) in cancerous tissues. We have examined the p-ERK in clinical samples (Figure 8A). The p-ERK expression was higher in tumor than the normal counter- parts, and the tendency was the same with BSSP4 expression. Increased BSSP4 expression and concomitantly elevated TR levels in HCC tissues of 12 representative paired HCC specimens are presented in Figure 8A. The correlation between TRs and BSSP4 was additionally analyzed. Linear regression analysis revealed a significant positive correlation between BSSP4 expression and TR levels, based on the T/N ratio (Spearman correlation coefficient = 0.445; 95% confidence interval (CI) = 0.154-0.665; P = 0.003) (Figure 8B). We have examined the correlation between BSSP4 and several parameters. A mild association between BSSP4 expression and distant metastasis is observed. Patients with T/N < 1.5 are less likely to have distant metastasis > 1 year than patients with T/N > 1.5 (P = 0.043). BSSP4 and VEGF levels in HCC tissues were determined using the Q-RT-PCR. We observed overexpression of BSSP4 and VEGF in 50% (26 of 52) and 61.5% (32 of 52) clinical specimens of HCC, respectively. Additionally, the T/N ratios of VEGF and BSSP4 expression revealed a significant positive correlation (Spearman correlation coefficient = 0.504; 95% confidence interval (CI) = 0.260-0.687; P = 0.0001) (Figure 8C; Table 1). We have added statistics information in “Additional file6: Figure S6” to show the correlation in BSSP4 expression and important parameters in HCC patients. Among them, bilirubin level was significantly (P = 0.043), cirrhosis was marginally (P = 0.057) associated with BSSP4 expression. Notably, the expression of BSSP4 between TRα (Additional file7: Figure S7A, Spearman r = 0.63, P < 0.001) and TRβ (Additional file7: Figure S7B, Spearman r = 0.6, P < 0.001) were demonstrated a significantly positive correlation in public Oncomine microarray data sets. Further, Kaplan-Meier analysis was used to analyze the correlation of BSSP4 expression with clinical parameters, including tumor size, tumor grade, tumor number, microvascular invasion and macrovascular invasion…etc. (Additional file8: Figure S8). Based on the data, the overall survival was significantly associated with HBsAg, ascites and tumor number. The HBsAg status was significantly associated with distant metastasis > 1y (p < 0.001). Moreover, the BSSP4 T/N ratio was also significatnly associated with distant metastasis > 1y (p = 0.043). The significantly associated parameters were shown in bold letters in the Additional file8: Figure S8. Finally, the significant (p < 0.01) correlation between TR and BSSP4 was found in clinical HCC specimens (Figure 8A, B, C). Thus, the TR and BSSP4 expression levels were positively associated in vitro or in vivo. Both our in vitro and in vivo findings support the use of BSSP4 as an effective therapeutic target for HCC therapy.

To determine the function of BSSP4 induced by T₃, J7-TRα1 cells were employed. BSSP4 mRNA and protein expression was regulated in a dose- and time-dependent manner in J7-TRα1 cells (Figure 8D). To further investigate whether the T₃-regulated effect in vitro is replicable in vivo, SCID mice were injected with J7-TRα1 cells and treated under several T₃ conditions[5]. Mice injected with J7-TR cells displayed multiple macroscopic lung tumor nodules, as determined by hematoxylin and eosin (H & E) staining. Tumor size was calculated per cm² of lung section and the metastatic index was significantly greater in hyperthyroid mice than in euthyroid condition. Higher T₃ levels (hyperthyroid conditions) enhanced BSSP4 expression and number of lung foci, as observed with IHC (Figure 8E, upper panel) and H & E staining (Figure 8E, lower panel)[5]. The BSSP4 expression level in metastatic lung section foci of hyperthyroid was obviously higher than euthyroid mice. Moreover, we observed T₃-induced cancer cell invasion and BSSP4 expression in vivo, suggesting that T₃ influences tumor motility via BSSP4 regulation.

Human hepatoma cells, HepG2, Huh7 and J7, were routinely cultured at 37°C in a humidified atmosphere of 95% air and 5% CO₂ in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS). HepG2 and J7 cell lines were stably transfected with TRα1 (HepG2-TRα1#1, HepG2-TRα1#2 and J7-TRα1) or TRβ1 (HepG2-TRβ1). The vector control cell line employed was HepG2-neo[42, 43]. Huh7-BSSP4 and J7-BSSP4 represent the Huh7 and J7 cell lines expressing BSSP4, respectively. Serum was depleted of T₃ (Td), as described previously[44].

HepG2-TRα1#1, HepG2-TRα1#2, HepG2-TRβ1, HepG2-neo, J7-TRα1, Huh7-BSSP4 and J7-BSSP4 cells were grown to confluence in 10 cm cell culture dishes. Cells contacting dishes were washed twice with PBS, and subsequently incubated in serum-free medium and either treated with T₃ for 24 h or left untreated. At the end of the treatment period, conditioned medium (CM) was collected and concentrated using spin columns with a molecular mass cut-off of 3 kDa (Amicon Ultra, Millipore, Billerica, MA).

Total cell lysates and conditioned media were prepared, and protein concentrations determined with the Bradford assay kit (Pierce Biotechnology, Rockford, IL). Equivalent amounts of proteins were fractionated on a 10% sodium dodecyl sulfate (SDS)-polyacrylamide gel. Separated proteins were transferred on to a nitrocellulose membrane (pH 7.9, Amersham Biosciences Inc., Piscataway, NJ), blocked with 5% non-fat powdered milk, incubated with specific primary antibodies at 4°C overnight, and subsequently hybridized with the respective secondary antibody (HRP-conjugated mouse/rabbit/goat anti-IgG) for 1 h at room temperature. Finally, immune complexes were visualized using the chemiluminescence method with an ECL detection kit (Amersham) on Fuji X-ray film, as described previously[45].

Total RNA was extracted from four T₃-treated HepG2 isogenic cell lines using TRIzol reagent, as described previously[46]. Subsequently, cDNA was synthesized via RT-PCR with the Superscript II kit (Life Technologies, Karlsruhe, Germany). Real-time Q-RT-PCR was performed on a 15 μl reaction mixture containing 750 nM forward and reverse primers, varying amounts of template and 1 × SYBR Green reaction mix (Applied Biosystems, Foster City, CA). SYBR Green fluorescence was determined using the ABI PRISM 7500 detection system (Applied Biosystems). Primers were designed using Primer Express Software (Applied Biosystems). Genes were normalized against the ribosomal binding protein (RiboL35A) gene. The human BSSP4 oligonucleotides used in this study include the forward primer, 5′-GGTCCCAGAAGGTGGGTGTT -3, and reverse primer, 5′- ACGCACCAGGGCAATGTC -3′.

Fragments of the BSSP4 promoter (positions -2066 to -7) were ligated into the pA3TK vector (Promega Corp., Madison, WI), based on the published sequence. Several serial deletion and mutant constructs of the BSSP4 promoter were amplified via PCR and cloned into pA3TK. Promoter sequences were confirmed using automated DNA sequencing. HepG2-TRα1 cells treated with 10 nM T₃ for 24 h were cotransfected with 0.6 μg DNA/well of pA3TK vector containing the BSSP4 promoter sequence and 0.3 μg of SVβ plasmid, a β-galactosidase expression vector (Clontech, Palo Alto, CA), in 24-well plates using TurboFect in vitro transfection reagent (Fermentas, Glen Burnie, MD) to determine the transcriptional activities of TREs within the BSSP4 promoter. At the end of the treatment period, transfected and non-transfected cells were lysed, and luciferase and β-galactosidase activities measured. Luciferase activity was normalized to that of β-galactosidase, as described earlier[47].

ChIP assays were performed to examine the interactions between TR and TRE on the BSSP4 promoter[45]. HepG2-TRα1 cells treated with 10 nM T₃ for 24 h or left untreated were harvested and cross-linked with 1% formaldehyde for 10 min at room temperature in DMEM medium. Reactions were terminated by adding 0.125 M glycine. Subsequently, cell lysates were washed three times with PBS, and resuspended in lysis buffer (150 mM NaCl, 5 mM EDTA, 50 mM Tris (pH 8.0), 0.1% SDS and 0.1% sodium deoxycholate) containing three protease inhibitors (1 mM PMSF, aprotinin, and leupeptin). Cell lysates were sonicated with a Misonix Sonicator 3000 Homogenizer (Mandel Scientific Company Inc., Guelph, ON, Canada) to disrupt chromatin. Sonicated DNA was between 200 and 1000 bp in length. Products were precleared with 60 μl protein A/G agarose (Sigma Chemicals, St. Louis, MO) for 2 h at 4°C. Complexes were immunoprecipitated with anti-TR (kindly provided by the laboratory of Dr. S-Y Cheng at the National Cancer Institute), anti-RXRα (Santa Cruz Biotechnology, Santa Cruz, CA) and anti-IgG antibodies (R & D Systems, Inc., Minneapolis, MN). The 100 bp fragment of the BSSP4 promoter containing the predicted TRE region was amplified via PCR with the forward primer, 5′- CTCCAGGAACGACAGGAGGGCG - 3′, and reverse primer, 5′-GCCTGGGTTTGGAGAGGCTGAAGTC- 3′.

Cells (4 × 10⁴) were seeded on a 6 cm dish and harvested at 1–7 days. The total number of cells in each condition provided a cell growth index via cell counting. Values are presented as fold increase in Huh7-BSSP4 and J7-BSSP4, relative to Huh7 and J7 control cells. Differences were analyzed using one-way ANOVA.

Total RNA (1 μg) was reverse-transcribed using Superscript II reverse transcriptase (Invitrogen) and Oligo (dT) to synthesize template cDNA. BSSP4 cDNA was amplified via PCR with the forward primer, 5S - CCAAGCTTATGGTGGTTTCTGGAGCGCCC-3′, and reverse primer, 5′ - CCGGAATTCCTAGGAGCGCGCGGCGG -3′, for 30 cycles at 95°C for 1 min, 60°C for 1 min and 72°C for 2 min. The BSSP4 open reading frame was ligated into pcDNA 3.0 expression vector, and the resulting construct sequenced to confirm the presence of the gene.

The Huh7 cell line was transfected with the BSSP4 cDNA construct in 10 cm cell culture dishes using Lipofectamine Reagent (Invitrogen). After 24 h, transfected cells were transferred to medium containing G418 (400 μg/ml) for selection until the generation of a single cell clone. Expression of BSSP4 in Huh7 and J7 cells was detected using Western blotting.

Short hairpin RNA clones targeting BSSP4 were purchased from the National RNAi Core Facility (Institute of Molecular Biology, Academia Sinica, Taiwan). Transfection of shRNA to against the endogenous BSSP4 gene in Mahlavu and J7-TR cells was transit performed using the Turbofect reagent (Invitrogen). The repression of the BSSP4 was confirmed by western blot analysis.

The influence of BSSP4 on migration and invasion abilities of Huh7-BSSP4 and J7-BSSP4 cells was determined with a rapid in vitro assay (Transwell) (Falcon BD, Franklin Lakes, New Jersey), as described previously[48]. Briefly, cell density was adjusted to 10⁵ cells/ml, and 100 μl of the suspension seeded on either non-matrigel-coated (migration) or matrigel-coated (invasion) (Becton-Dickinson) upper chambers of the Transwell plate. For both assays, the pore size of the upper chamber was 8 mm. The medium in the upper chamber was serum-free DMEM, while the lower chamber contained DMEM supplemented with 20% fetal bovine serum (FBS). After incubation for 24 h at 37°C, cells traversing the filter from the upper to lower chamber were examined via crystal violet staining and cell counting. Experiments were performed at least three times.

Formalin-fixed and paraffin-embedded tissues from the lung of SCID mice were evaluated by hematoxylin and eosin (H & E) staining and immunohistochemistry using polyclonal antibody against BSSP4 (GeneTex, Inc, San Antonio, Texas) after the avidin-biotin complex method, as described previously. The positive staining consisted of cancer cells with dark brown of BSSP4 immunoreactivity.

Similar conditions were employed with SCID mice containing various T₃ levels (Group A to B) induced via injection of J7-TR cells[5]. Mice were divided into two groups, specifically, Group A (euthyroid) comprising control mice given normal drinking water, and Group B (hyperthyroid) administered drinking water augmented with T₃ (2 mg/L) (Sigma Chem. Co., St. Louis, MO) after inoculation of tumor cells. Mice were sacrificed after about 1 month injection, livers and lungs were removed for tumor biopsy, and the T₃ and TSH levels determined. The T₃ and TSH levels in the serum of group A were 45.5 ng/dl and 0.246 mIU/ml, and in group B were 619 ng/dl and 0.008 mIU/ml, respectively. Tumor volume was calculated using the following equation: length × height × width. Formalin-fixed and paraffin-embedded tissues from SCID mice were evaluated by hematoxylin and eosin (H & E) staining and immunohistochemistry using polyclonal antibody against BSSP4 (GeneTex). All procedures were performed under sterile conditions in a laminar flow hood. Animal experiments were performed in accordance with United States National Institutes of Health guidelines and Chang-Gung Institutional Animal Care and Use Committee Guide for the Care and Use of Laboratory Animals.

With informed consent, patients with HCC diagnosed between 2000 and 2003 were selected consecutively for this study. All samples of HCC tissues with paired adjacent normal liver tissues were obtained during surgical resection from the Chang Gung Memorial Hospital medical research center for western blot and Q-RT-PCR analysis. The study protocol was approved by the Medical Ethics and Human Clinical Trial Committee at Chang-Gung Memorial Hospital.

Data are expressed as mean values ± SEM of at least three experiments. Statistical analysis was performed using the Student’s t test and one-way ANOVA analysis. P < 0.05 was considered statistically significant.