Notice: Wiley Online Library will be unavailable on Saturday 30th July 2016 from 08:00-11:00 BST / 03:00-06:00 EST / 15:00-18:00 SGT for essential maintenance. Apologies for the inconvenience.
Potential conflict of interest: Nothing to report.
The study was funded by the Hong Kong Research Fund for the Control of Infectious Diseases (05050022), the Hong Kong Research Grants Council Collaborative Research Fund (HKU 1/06C and 7/CRF/09), and the Hong Kong University Small Project Fund (200707176172). I.O.L.N. is Loke Yew Professor in Pathology.
Random integration of hepatitis B virus (HBV) DNA into the host genome is frequent in human hepatocellular carcinoma (HCC) and this leads to truncation of the HBV DNA, particularly at the C-terminal end of the HBV X protein (HBx). In this study, we investigated the frequency of this natural C-terminal truncation of HBx in human HCCs and its functional significance. In 50 HBV-positive patients with HCC, full-length HBx was detected in all nontumorous livers. However, full-length HBx was found in only 27 (54%) of the HCC tumors, whereas natural carboxylic acid (COOH)-truncated HBx was found in the remaining 23 (46%) tumors. Upon clinicopathological analysis, the presence of natural COOH-truncated HBx significantly correlated with the presence of venous invasion, a hallmark of metastasis (P = 0.005). Inducible stable expression of the COOH-truncated HBx protein (with 24 amino acids truncated at the C-terminal end) enhanced the cell-invasive ability of HepG2 cells, as compared to full-length HBx, using the Matrigel cell-invasion assay. It also resulted in increased C-Jun transcriptional activity and enhanced transcription of matrix metalloproteinase 10 (MMP10), whereas activation of the MMP10 promoter by COOH-truncated HBx was abolished when the activator protein 1–binding sites on the MMP10 promoter were mutated. Furthermore, silencing of MMP10 by short interfering RNA in HBxΔC1-expressing HepG2 cells resulted in significant reduction of cell invasiveness. Conclusions: Our data suggest that COOH truncation of HBx, particularly with 24 amino acids truncated at the C-terminal end, plays a role in enhancing cell invasiveness and metastasis in HCC by activating MMP10 through C-Jun. (HEPATOLOGY 2013)
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Hepatocellular carcinoma (HCC) is one of the major malignancies worldwide and the second-most common fatal cancer in Southeast Asia, China, and Hong Kong, as a result of the high prevalence of hepatitis B virus (HBV) infection. HBV is a partial double-stranded DNA virus with a 3.2-kb genome containing four open reading frames, including the viral DNA polymerase (P), viral envelope (surface antigens) proteins (PreS1, PreS2, or S), core proteins (PreC or C), and HBV X protein (HBx). Integration of the HBV DNA into the host genome is common in HCC and this may lead to alterations of the host cells by disrupting the expression of cellular genes that are important for cell growth, survival, and cellular differentiation. These cellular genes include cyclin A2,1 retinoic acid receptor,2 and human telomerase reverse transcriptase (hTERT).3, 4 Moreover, full-length HBx can alter the expression of cellular genes by transcription factors, including nuclear factor kappa B (NF-κB), activator protein 1 (AP-1), cyclic adenosine monophospahte response element-binding protein (CREB), and TATA-binding protein (TBP), and can promote cell survival.5
It is well established that random HBV genome integration can lead to truncation of the HBV genome, especially on the HBx gene locus at the C-terminus.6-8 Furthermore, ectopic expression of the truncated, but not full-length, form of HBx leads to overgrowth of tumor cells in mouse models.6, 7 Previous studies have observed enhanced cell invasiveness with full-length HBx in in vitro studies; however, the effects of HBx with C-terminal truncation remain to be investigated.9-11
In this study, by examining the status of HBx integration in human HCC samples, we found a significant association between the presence of C-terminal truncation of HBx DNA and venous invasion. Furthermore, ectopic expression of the carboxylic acid (COOH)-truncated HBx, particularly with the breakpoint at 130 amino acids (aa), enhanced cell invasiveness of HCC cells in vitro by activation of C-Jun/AP-1 and increased the transcription of matrix metalloproteinase protein 10 (MMP10). Our data suggest that COOH truncation of HBx may play a role in enhancing cell invasiveness and metastasis in human HCC.
Fifty pairs of human HCCs and their corresponding nontumorous liver tissues from patients with liver resection for HCC between 1992 and 2001 at Queen Mary Hospital, Hong Kong, were randomly selected for study. These 50 patients had positive serum hepatitis B surface antigen (HBsAg) status. patients' ages ranged from 34 to 70 years; 43 were male and 8 female. All specimens were snap-frozen in liquid nitrogen and kept at −80°C. Frozen sections were cut from nontumorous liver and tumor blocks separately and stained for histological examination to ensure a homogenous cell population of tissues. Use of human samples was approved by the institutional review board of the University of Hong Kong/Hospital Authority Hong Kong West Cluster.
HCC cell lines including HepG2, Hep3B, PLC/PRF/5, HLE, Huh7, BEL7402, and SMMC7721 and an immortalized normal liver cell line LO2 were maintained in Dulbecco's modified Eagle minimal essential medium (DMEM) high glucose (GIBCO-BRL, Grand Island, NY), supplemented with 10% fetal bovine serum (FBS). The other two HCC cell lines, SNU182 and SNU449, were grown in RPMI-1640 medium (GIBCO-BRL), supplemented with 1 mM of sodium pyruvate and 10% FBS. The other immortalized healthy liver cell line (MIHA) was maintained in DMEM high glucose, supplemented with 10% FBS and 1 mM of sodium pyruvate.
Full-length HBx DNA (ayw subtype; GenBank no.: U95551) was amplified from the HBx/pcDNA3.1+ plasmid10 and subcloned into Myc/pLVX-Tight Puro and Myc/pcDNA3.1+ vectors. HBx truncation mutant (named HBxΔC1) with 24 aa of HBx was made and subcloned into Myc/pLVX-Tight Puro and Myc/pcDNA3.1+ vectors. In addition, wild-type (WT; −1,077 to +1) MMP10 promoter was amplified from healthy human liver DNA. A mutant with mutations at the putative AP-1 transcription factor binding sites on the MMP10 promoter was subcloned into pGL3-Basic vector for reporter assay. Moreover, the TATA box from the herpes simplex virus thymidine kinase promoter from Clontech (pTA; Clontech Laboratories, Inc., Palo Alto, CA) and six times repeat of AP-1-binding site sequence (5′-TGACTAA-3′) fused with pTA promoter (6XAP-1) were subcloned into pGL3-Basic vector for reporter assay.
DNA and RNA Extraction and Semiquantitative Reverse-Transcriptase Polymerase Chain Reaction.
Total DNA was isolated from the 50 pairs of HCCs and their corresponding nontumorous liver tissues according to the standard protocol, as described previously.12 Total RNA of 11 hepatoma cell lines was extracted using TRIzol (Invitrogen, Carlsbad, CA), according to manufacturer's protocol. For polymerase chain reaction (PCR) amplification of HBx, sets of PCR primers (44F: 5′-TCCTTTGTTTACGTCCCGTC-3′, 197R:-5′GCAGATGAGAAGGCACAGAC-3′ and 465R: 5′-TTAGGCAGAGGTGAAAAAGTTGC-3′) were used for full-length and COOH-truncated HBx, respectively (Fig. 1A). In addition, to detect the presence of truncation at 130, 140, and 150aa of COOH-truncated HBx, respectively, sets of PCR primers (1F: 5′-ATGGCTGCTAGGCTGTGCT-3′, 390R: 5′-ATCTAATCTCCTCCCC-3′, 420R: 5′-CAATTTATGCCTACAGCCTCCTAC-3′ and 450R: 5′-TTAGTTGCATGGTGCTGGTGCGCAG-3′) were used (Supporting Fig. 1A).
A set of PCR primers (5′-ATCCAGTTTGGTGTCGCGGAGC-3′ and 5′-GAAGGGGAAGACGCACAGCT-3′) was used to amplify MMP10 complementary DNA (cDNA), with β-actin (primer set of 5′-GTCACTTCAGCTCCTTTCCT-3′ and 5′-ATCTTGCGAAAGGCGGAACT-3′) used as a reference for the amount of cDNA added in the PCR reactions. The detailed protocol for HBx-specific Alu-PCR was according to that described previously by Minami et al.13 Primers used for HBx-specific Alu-PCR were according to sequences described by Murakami et. al.14 Amplified PCR products were subjected to DNA sequencing.
Immunohistochemistry (IHC) was performed on formalin-fixed, paraffin-embedded sections as previously described,10 using rabbit polyclonal antibody (Ab) against HBx (a gift by Dr. MA Feitelson) at 1:5,000 dilution.
Establishment of Tetracycline-Inducible HBx-Expressing Cells.
The HepG2 cell line was, first, transfected with pLVX Tet-Off Advanced vector (Clontech Laboratories, Inc., Mountain View, CA) using Lipofectamine 2000 (Invitrogen), according to manufacturer's protocol. tTA(Tet-Off)-expressing cells were selected with G418 at 1 mg/mL for 14 days. To obtain stable inducible HBx-expressing cells, lentivirus containing full-length and C-terminal truncated HBx in Myc/pLVX-Tight Puro vector was infected into tTA-expressing HepG2 cells and selected with puromycin at 1 μg/mL for 7 days.
Cell-invasion assay was performed with Matrigel precoated transwell chamber (BD Biosciences, Sparks, MD). Cells (3 × 105) of cells were seeded onto the transwell chamber and were allowed to invade through the extracellular matrix to the lower chamber. Invaded cells were fixed with 3.7% formaldehyde and stained with crystal violet. Three randomly selected fields on the fixed transwell chamber were captured by photography, and invaded cells were counted. The experiment was performed at least thrice independently.
Western Blotting Analysis.
Cells were lyzed in sodium dodecyl sulfate (SDS) containing buffer and equal amounts of protein were separated in SDS/polyacrylamide gel electrophoresis gel for western blotting analysis. Immunodetection was performed using anti-Myc (Santa Cruz Biotechnology, Santa Cruz, CA), anti-α-tubulin (Sigma-Aldrich, St. Louis, MO), and anti-C-Jun (BD Biosciences, Franklin Lakes, NJ) Abs.
Dual Luciferase Reporter Assay.
HepG2 cells were transfected with different combinations of plasmids using FuGENE 6 reagent (Roche, Indianapolis, IN), according to the manufacturer's protocol. Plasmids used included Myc/pcDNA3.1+ vector containing various forms of HBx, MMP10-WT/pGL3-Basic, MMP10-AP1-Mut/pGL3-Basic reporter constructs, and an internal control (pRL-SV40). The total amount of expression vectors was equalized with the empty vector. Twenty-four hours after transfection, luciferase and Renilla luciferase activities were measured by the Dual Luciferase Reporter assay system (Promega, Madison, WI), according to the manufacturer's protocol. Transfection efficiency was normalized with the Renilla luciferase activity. Experiments were done thrice independently.
Chromatin Immunoprecipitation Assay.
Cells (3 × 106) were seeded 1 day before harvest and chromatin immunoprecipitation (ChIP) assay was performed. Cells were fixed with 1% formaldehyde for 10 minutes, and the reaction was neutralized by adding glycine to a final concentration of 125 mM in the mixture. Formaldehyde cross-linked cells were collected by centrifugation, resuspended in membrane containing lysis buffer (5 mM of KOH [pH 8.0], 85 mM of KCL, 0.5% NP-40, 0.5% SDS, and 1×CompleteProtease Inhibitors), and incubated on ice for 30 minutes. Cell nuclei were collected by centrifugation, and cross-linked DNA was digested by Micrococcal nuclease for 20 minutes, according to manufacturer's protocol (New England Biolabs, Inc., Ipswich, MA). Digested DNA was released from nuclei by freeze-thaw cycles and processed for ChIP assay according to the EZ-Chip assay kit (Millipore, Billerica, MA) protocol. The Ab against C-Jun protein was used (Santa Cruz Biotechnology), and the primer set (forward 5′-CAAACACAGAAATCATTTCCTGG-3′ and reverse 5′-AGATCACCAACAGTATGATTCATGC-3′) covering the putative AP-1-binding site on the MMP10 promoter was employed for standard PCR measurement in the ChIP assay.
Clinicopathological Correlation and Statistical Analysis.
Clinicopathological features of HCC patients, including tumor size, cellular differentiation according to Edmondson's grading, venous invasion into portal or hepatic venules, direct liver invasion, tumor microsatellite formation, tumor encapsulation, and number of tumor nodules, were analyzed using PASW Statistics 18 for Windows (SPSS, Inc., Chicago, IL). For clinicopathological correlation analysis, Fisher's exact test was used for analysis of categorical data. For in vitro cell-invasion assay and reporter assay, the Student t test was used for continuous data. Results were considered significant if the P value was less than 0.05.
Presence of Natural C-Terminal-Truncated HBx Mutant in Human HCCs and Its Association With Venous Invasion.
All of the 50 patients who were serologically positive for HBsAg had HBx in their livers (Fig. 1A). In all these 50 HBx-positive patients, full-length HBx was detected in nontumorous liver tissues, using PCR primers that flanked the C-terminal end of full-length HBx DNA (Fig. 1A). Interestingly, full-length HBx was detected in only 27 (54.0%) of these 50 tumors. However, in the remaining 23 (46.0%) HCCs without the full-length HBx in the tumors, the N-terminal HBx DNA fragment was detected upon PCR using another reverse PCR primer flanking the 197 nucleotides (nt) of HBx, indicating the presence of C-terminal-truncated HBx (Fig. 1A). Furthermore, the breakpoint between 125 and 135 aa was the major form of truncation, being detected in 11 (47.8%) of the 23 cases (Supporting Fig. 1). In the 23 cases showing COOH-truncated HBx messenger RNA (mRNA) expression, 22 (95.6%) showed positive HBx immunostaining (Supporting Fig. 2).
Upon clinicopathological correlation, we found that patients with C-terminal-truncated HBx in their tumor tissues had significantly more venous invasion, a feature of metastasis (P = 0.005) (Table 1). There was no significant correlation between the presence of C-terminal-truncated HBx in tumors and the remaining pathological features (Table 1).
Table 1. Clinicopathological Correlation of HBx COOH Truncation in Human HCCs
Cellular differentiation according to Edmondson's grading
Background liver diseases
Normal and chronic hepatitis
Detection of Natural C-Terminal-Truncated HBx in HCC Cell Lines.
We also analyzed the expression status of HBx and the presence of the HBx truncated forms in HCC cell lines by reverse-transcriptase (RT)-PCR using the primer pair flanking the C-terminal end of full-length HBx (Fig. 1A). Of the nine HCC cell lines and the two immortalized healthy liver cell lines (LO2 and MIHA) tested, only the PLC/PRF/5 cell line was found to express the full-length HBx transcript (Fig. 1B). A small amount of N-terminal end, but not full-length, HBx mRNA was detected in the Hep3B cell line using the reverse primer with flanking 197 nt of HBx (Fig. 1B). This indicates that full-length HBx is expressed in PLC/PRF/5 cells and that C-terminal-deleted HBx is expressed in Hep3B cells.
Stable Expression of COOH-Truncated HBx Protein Enhanced Invasiveness of HCC Cells.
To delineate the mechanistic basis of our observed association between natural COOH-truncated HBx and venous invasion in human HCC samples, to this end, we performed the in vitro cell-invasion assy. To compare the effect on cell-invasion ability among the various forms of HBx in HCC cells, the tetracycline/doxycycline inducible expression system (Tet-Off system) was successfully generated and employed to express the full-length and COOH-truncated form of HBx, respectively. For the COOH-truncated form of HBx, we chose the one with a breakpoint at 130 aa (HBxΔC1)6, 8, 15 (Fig. 2A), which was previously reported and was also the major form of COOH-truncated HBx in our human HCCs (Supporting Fig. 1) for further studies. Interestingly, in the cell-invasion assay, induced stable expression of both full-length and COOH-truncated HBx (HBxΔC1) significantly enhanced the invasiveness of HepG2 cells, as compared to the corresponding vector control. The highest number of invaded cells was observed in the COOH-truncated HBxΔC1-expressing cells, as compared to either the vector control or full-length HBx (Fig. 2B). Furthermore, we also established another tetracycline/doxycycline-inducible expression system (Tet-On system) and used it to express the full-length and COOH-truncated form of HBx. The result consistently showed that the COOH-truncated form of HBx had enhanced cell invasiveness, as compared with the full-length form HBx expressing Tet-On HepG2 cells (Supporting Fig. 3A). Although ectopic expression of HBxΔC1 in the healthy liver cell line, LO2, lost the growth-suppressive effect of the full-length HBx, as shown with the colony-formation assay (CFA) (Supporting Fig. 4A), mild expression of full-length or COOH-truncated HBx did not show a significant alteration of cell-proliferation rates in the inducible HBx-expressing HepG2 cells (Supporting Fig. 4B).
Transcriptional Activation of MMP10 in HBxΔC1-Expressing Cells.
Transcription of the MMP family may be regulated by the Ras/Raf/MEK/ERK-signaling cascade, and Erk phosphorylation may lead to transcription activation of MMP9 in HBx-transfected cells.9, 16 Therefore, we queried whether COOH-truncated HBx could activate the MMP family in HCC cells. With semiquantitative RT-PCR, we observed that stable expression of HBxΔC1 increased MMP10 mRNA levels in Tet-Off HepG2 cells, as compared to full-length HBx and vector control (Fig. 3A). In addition, MMP10 mRNA transcripts were also increased in Tet-On HBxΔC1-expressing HepG2 cells, as compared with full-length HBx-expressing cells (Supporting Fig. 3B).
We then queried whether AP-1 transcription factor binding sites were involved in the activation of MMP10 transcription by COOH-truncated HBx protein. Dual luciferase reporter assay was performed with either WT (MMP10 WT) construct or MMP10 promoter construct, but with mutations of the putative AP-1-binding sites (MMP10 AP-1 Mut). There was a 1.8-fold induction of WT MMP10 promoter activity when HBxΔC1 was coexpressed, as compared to the vector control and full-length HBx (Fig. 3B). However, when the AP-1-binding sites were mutated in COOH-truncated HBx-expressing cells, MMP10 promoter activity was reduced by 71%, suggesting that COOH-truncated HBx induced MMP10 mRNA expression by AP-1 activation (Fig. 3B). Furthermore, with silencing of MMP10 by short interfering RNA (siRNA) in HBxΔC1-expressing HepG2 cells, there was a 75% reduction of cell invasiveness, as compared to nontarget control transfected cells, suggesting that HBxΔC1 enhanced cell-invasive ability by MMP10 (Supporting Fig. 5).
Increased C-Jun Transcriptional Activity and Protein Expression in HBxΔC1-Expressing Cells.
It is well documented that the Jun and Fos transcription factor protein family is involved in transcription by the AP-1 transcription factor binding site.17 To further assess the transcriptional activity of AP-1 transcription factors, we transiently transfected pGL3-basic vector containing six repeats of AP-1 consensus binding sequence and its corresponding vector only in the Tet-Off HepG2 cell system. There was a 1.5-fold induction of 6xAP-1 promoter activity when HBxΔC1 was coexpressed, as compared to the vector control and full-length HBx, suggesting that HBxΔC1 induced AP-1 transcriptional activity in HepG2 cells (Fig. 4A). Moreover, a previous report has suggested that C-Jun protein is overexpressed in human HCC, as assessed by IHC.18 Therefore, we examined the protein expression of C-Jun in full-length HBx- and HBxΔC1-expressing HepG2 cells. There was an observable induction of C-Jun protein in HBxΔC1-expressing cells, as compared to full-length HBx-expressing or vector control cells (Fig. 4B). Similarly, C-Jun protein was overexpressed in HBxΔC1-expressing cells, as compared to full-length HBx expressing cells in the Tet-On HepG2 cell system (Supporting Fig. 3C). Furthermore, with the ChIP assay, a significant amount of C-Jun protein interacted with the MMP10 promoter in HBxΔC1-expressing Tet-Off HepG2 cells, as compared to full-length HBx-expressing or vector control cells (Fig. 4C). Taken together, these results indicate that HBxΔC1 is able to increase both C-Jun protein expression and transcriptional activity, resulting in enhanced MMP10 transcription in HepG2 cells.
Evidence from previous studies suggests that HBV genomic DNA integration or mutation leads to COOH truncation of the HBx protein in human HCC.6-8, 19 However, such integration or mutation is uncommon in corresponding nontumorous liver tissues. In our present study, 46% (23 of 50) of human HCC tissues contained COOH-truncated HBx DNA. The result was consistent with that of a recent study showing that 79% of human HCCs from China had COOH-truncated HBx transcript in tumor tissues.7 These lines of evidence indicate that COOH-truncation of HBx is frequent in HCC. Furthermore, upon clinicopathological correlation, we found that the presence of COOH-truncated HBx in HCC tumors was associated with venous invasion. So, the presence of COOH-truncated HBx appears to have clinical significance.
Because of the growth-suppressive and toxic effects of HBx on host cells, it has been difficult to establish a stable cell line with HBx expression.20 In the present study, we successfully established the Tet-Off HBx expression system in HepG2 cells that efficiently and effectively allowed controlled expression of HBx. Such inducible systems have also been used by other groups, but they worked only with full-length HBx and not on COOH-truncated HBx.21, 22 We then attempted to delineate the mechanistic basis of our observed association between the presence of natural COOH-truncated HBx and venous invasion in human HCCs by assessing cell-invasive ability in vitro. We chose the previously widely reported COOH-truncated form of HBx, with a breakpoint at 130 aa (HBxΔC1)6, 8, 15 for our cell model because breakpoint between 125 and 135 aa was the major form of truncation (47.8% of the 23 cases) (Supporting Fig. 1). With the cell-invasion assay, we observed enhanced HepG2 cell invasiveness with both full-length HBx and HBxΔC1, but more so with HBxΔC1 in the inducible expression system. In addition, in the transient expression of nontagged full-length and COOH-truncated HBx, only the HBxΔC1-expressing cells showed remarkably enhanced cell invasiveness, as compared to full-length HBx-expressing or empty vector control cells (Supporting Fig. 6), consistently indicating that HBxΔC1 is more potent in enhancing cell invasiveness of HCC cells.
To further dissect the mechanistic basis of the HBxΔC1-induced cell invasiveness, we observed an increased C-Jun expression as well as activation in the HBxΔC1-expressing HepG2 cells, as compared to the vector control and full-length HBx. Along with increased C-Jun activation, there was up-regulation in MMP10 transcription. Furthermore, the increased promoter activity of WT MMP10 by HBxΔC1 was abolished by mutating the AP-1 sites of the MMP10 promoter. This suggests that HBxΔC1 activates C-Jun signaling, which, in turn, up-regulates MMP10 transcription, as shown with the ChIP assay. Additionally, silencing of MMP10 by siRNA in HBxΔC1-expressing HepG2 cells resulted in a significant reduction of cell invasiveness, suggesting that HBxΔC1 enhanced cell-invasive ability by MMP10. However, the same reasoning may not be applicable to explain the enhanced cell invasiveness induced by full-length HBx (Fig. 2), because MMP10 transcriptional up-regulation by C-Jun activation was not observed with full-length HBx. In our previous study, ectopic expression of full-length HBx could up-regulate the transcription of another invasiveness-related gene, urokinase-type plasminogen activator, by activation of NF-κB.10 Moreover, several reports have shown that HBx can lead to up-regulation of other MMP protein family members, such as MMP1, MMP2, MMP3, and MMP9,9, 11, 23-25 metastasis-associated protein 1, and histone deacetylase 1,26 suggesting that other mechanisms may contribute to enhanced cell invasion induced by full-length HBx. Nevertheless, there was a slight induction of AP-1-mutated MMP10 promoter activity in HBxΔC1-expressing cells, as compared to full-length HBx-expressing or vector control cells (Fig. 3B). Such an observation implies that additional transcription factor activation might be involved with HBxΔC1 in HepG2 cells, and further studies are warranted.
Previous studies have shown that both natural COOH-truncated HBx and HBx with point mutation at the C-terminus enhanced HCC cell growth, as compared to full-length form of HBx, resulted in formation of larger tumors in vivo.6, 7, 20 In the present study, we observed that HBxΔC1 lost the growth-suppressive effect of full-length HBx as, shown by CFA in vitro (Supporting Fig. 4A). However, we did not observe an association between the presence of COOH-truncated HBx and tumor size in human HCCs in this study. It is to be noted that the gene loci of HBV integration, single-nucleotide polymorphism, and point mutations of HBx27 can be factors contributing to the reduction of the antiproliferative ability of full-length HBx and perhaps HCC tumor size in patients. On the other hand, mild expression of full-length HBx did not significantly alter cell-proliferation rates, as compared to vector control and HBxΔC1-expressing HepG2 cells (Supporting Fig. 4B), suggesting that the expression level of full-length HBx may be important in relation to antiproliferative function in HCC cells. Further investigation is needed.
Recent focus has been placed on the importance of HBV integration in HCC tumor samples. It has been found that the breakpoint within the HBV genome is usually at the C-terminus of HBx at approximately 1,800 base pairs.28 The result is consistent with our current PCR-based study of COOH-truncated HBx in human HCC; however, the integrated sites of HBV DNA into the host genome in our HCC tumors varied (Supporting Fig. 1C), suggesting that the integration sites may not be directly associated with effects on cell invasiveness in human HCC.
Although full-length HBx is less potent in enhancing the cell invasion of HCC cells, 54% of our human HCCs had full-length HBx. In various previous studies, it has been shown that full-length HBx could induce tumor formation in transgenic mice or increase susceptibility to carcinogen-induced hepatocarcinogenesis,29-31 suggesting that full-length HBx may play an important role in tumor initiation.
To conclude, our data suggest that COOH truncation of HBx enhances the cell invasiveness of HCC cells in vitro and is associated with venous invasion in HCC patients. Our data also suggest that COOH-truncated HBx, particularly with the breakpoint at 130 aa, induces MMP10 transcription by C-Jun/AP-1 activation. Taken together, COOH truncation of HBx in human HCC may play a significant role in enhancing cell invasiveness and cancer metastasis.