To facilitate the introduction of the HBx gene into primary hepatocytes, a recombinant baculoviral vector was constructed using pAcSG2-CMV, which contains the eukaryotic gene expression cassette derived from pIRES-EGFP (Clontech, Mountain View, CA) (23). The gene sequences for the enhanced green fluorescent protein and internal ribosome entry site were removed using BamHI and NotI. The DNA fragment coding HBx was amplified using the primers 5'-CTAGCTAGCATGGCTGCTCGGGTGTG-3' and 5'-AACTGCAGTTAGGCAGAGGTGAAAAAGTTGC-3', and using pGEX-4T-HBx (24) as a template. The PCR product was introduced into pAcSG2-CMV downstream of the cytomegalovirus promoter and was confirmed by sequencing. Recombinant baculoviruses were produced in Sf9 cells (BD Biosciences Pharmingen, San Diego, CA) cotransfected with baculoviral Gold DNA (BD Biosciences) and baculoviral transfer vectors. Recombinant baculoviruses were amplified in Sf9 cells and concentrated by centrifugation of the culture supernatant at 6,000 rpm for 16 h at 4℃. The virus pellet was resuspended in DMEM-F12 medium (GibcoBRL, Carlsbad, CA) and the viral titer was determined using a plaque forming unit assay.

Isolation of hepatocytes was done as described previously (23). Briefly, the livers of male C57BL/6 (B6) mice (Deahan Biolink, Seoul, Korea) were perfused with liver perfusion medium (GibcoBRL) via the portal vein. Released hepatocytes were washed, seeded into 48-well plates at a density of 2×10⁴ hepatocytes/well and allowed to adhere overnight. Recombinant baculoviruses were added to the adhered hepatocyte culture and 1 h later the culture medium was replaced.

Lymph nodes and spleen cells were isolated from MataHari TCR transgenic mice (25) that were generously provided by Dr. Matzinger, United States National Institutes of Health. These mice have CD8 T cells expressing TCR specific to the H-Yp^b peptide bound to H-2D^b. CD8⁺ T cells were purified using magnetic beads by incubation with beads coupled to anti-mouse CD8 antibody (Miltenyi Biotec, Auburn, CA) for 30 min at 4℃ prior to passage twice through a MACS magnetic cell separation column (Miltenyi Biotec). CD8⁺ cell purity was ascertained by labeling cells with PE-conjugated anti- CD8 antibody (BD Biosciences) and flow cytometry using the BD Vantage system (BD Biosciences). The purity was consistently 90~98%.

Purified MataHari CD8⁺ T cells (2×10⁵ cells/well) were added to hepatocyte cultures pre-infected with baculoviruses. To analyze the proliferation of CD8 T cells, the cells were labeled with CFSE (Molecular Probes, Eugene, OR) prior to addition to the hepatocyte-culture. After 2 days, CD8 T cells were harvested in phosphate-buffered saline containing 1 mM EDTA, stained with antibodies against CD8 and CD45 (BD Biosciences) and analyzed using flow cytometry (Becton Dickinson, Sunnyvale, CA). To enumerate apoptotic cells, unlabeled CD8 T cells were co-cultured, harvested and stained with antibodies against CD8 and CD45, and then a terminal deoxynucleotidyltransferase-mediated dUTP-biotin nick end-labeling (TUNEL) assay was conducted (Roche Applied Science, Indianapolis, IN).

For the production of IFN-γ and IL-10 in co-culture of CD8 T cells and primary hepatocytes, the supernatant obtained from a 48 h culture was used for an enzyme-linked immunosorbant assay (ELISA) using an IFN-γ ELISA kit (Pierce Endogen, Rockford, IL) and an IL-10 ELISA kit (Biosource, Carlsbad, CA).

For detection of HBx transcript in baculovirus infected hepatocytes, total RNA of hepatocytes infected with recombinant baculoviruses was isolated at the indicated time points using RNAzol-B (Tel-Test, Friendswood, TX). To remove contaminated vector DNA, the RNA samples were treated with 1 U/µl DNase I (Invitrogen) for 15 min at room temperature and were reverse transcribed using oligo-dT primers. PCR was performed using cDNA or RNA as a template and a primer set for HBx.

For the analysis of transcriptional levels of H-2K, ICAM-1 or PD-L1, total RNA of hepatocytes from male C57BL/6 or HBx transgenic mice (26) (kindly provided by Dr. Dae-Yul Yu, KRIBB, Korea) was isolated. The cDNA was synthesized using the Superscript II first-strand cDNA synthesis system (Invitrogen) according to the manufacturer's instructions. Semi-quantitative RT-PCR was performed using serially diluted cDNA and primer sets for H-2k, ICAM-1 or PD-L1: H-2k forward, 5'-CTGCAGGGGATGGAACCTTC, H-2k reverse, 5'-CTTCACGCTAGAGAATGAGG; ICAM-1 forward, 5'-TGGAACTGCACGTGCTGTAT, ICAM-1 reverse, 5'-ACCATTCTGTTCAAAAGCAG, PD-L1 forward, 5'-GGAATTGTCTCAGAATGGTC, PD-L1 reverse, 5'-GTAGTTGCTTCTAGGAAGGAG. For normalization of PCR product, RT-PCR for beta-actin was conducted using forward primer, 5'-CAGGTCATCACCATTGGCAATGAG and reverse primer, 5'-CAGCACTGTGTTGGCGTACAGGTC. PCR products were separated using a 1.5% agarose gel and band intensities were analyzed with the Molecular Imager Gel Doc XR System (Bio-Rad Laboratories, Hercules, CA). For detection of HBx protein, HEK293 cells transfected with baculovirus transfer vector and Huh7 cells infected with recombinant baculoviruses were lysed in RIPA buffer containing 1 mM 4-(2-aminoethyl)-benzenesulfonyl fluoride, 0.3 µM aprotinin and 1 µM leupeptin. Proteins were separated by 15% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to a polyvinylidene difluoride membrane. HBx protein was visualized by Western blot analysis using rabbit sera specific to HBx (24) and the enhanced chemiluminescent system (GE Healthcare, Buckinghamshire, UK).

For efficient introduction of the HBx gene into primary hepatocytes, recombinant baculoviruses containing the gene in the eukaryotic gene expression cassette were produced and utilized. HBx gene expression in mammalian cells using the baculoviral vector system was demonstrated (Fig. 1). Using an antibody specific to HBx, a protein with an approximate molecular weight of 17 kDa was detected in Western blots of HEK293 cells transfected with the baculoviral transfer vector harboring the HBx gene (pAcSG2-CMV-HBx) (Fig. 1A). Western blots of HEK293 cells transfected with the control empty vector pAcSG2-CMV did not detect the 17 kDa band (Fig. 1A). HBx was also detected in Huh7 human hepatoma cells infected with recombinant baculoviruses containing the HBx gene, but not in Huh7 cells infected with control baculoviruses lacking the HBx gene (Fig. 1B). HBx expression was also evident in primary murine hepatocytes infected with recombinant baculoviruses using reverse transcription-polymerase chain reaction (RT-PCR) (Fig. 1C). HBx gene transcription was detected in hepatocytes infected with baculoviruses containing the HBx gene for 3 days post-infection but not with control baculoviruses. The possibility of amplification of HBx gene from transduced baculoviral DNA was excluded by the absence of the gene in PCR using RNA samples as templates.

The influence of HBx expression in hepatocytes on apoptosis of CD8 T cells recognizing hepatocyte surface antigens was assessed next. For this, apoptosis of CD8⁺ T cells cultured with hepatocytes that did and did not express HBx was determined. CD8 T cells were isolated from female MataHari mice, which express a transgenic T-cell receptor (TCR) specific to the male antigenic peptide H-Yp^b presented by H-2D^b. Hepatocytes were isolated from male C57BL/6 (H-2^b) mice and infected with baculoviruses containing the HBx gene prior to co-culture with CD8 T cells. As a result, hepatocytes presented antigen to MataHari CD8 T cells in co-culture and expressed HBx when these cells encountered CD8 T cells. After 48 h of co-culture, the percentage of apoptotic CD8 T cells in the presence of expressed HBx was higher than that of the two controls (cultures of non-infected hepatocytes and hepatocytes infected with control baculoviruses), while the controls were similar to one another (Fig. 2). These results suggest that HBx expression may render hepatocytes capable of promoting apoptosis of CD8 T cells that recognize hepatocyte surface antigen.

To assess whether the increased frequency of CD8 T cell apoptosis was due to aberrant activation of CD8 T cells by HBx-expressing hepatocytes, CD8 T cell proliferation in co-culture was analyzed. CD8⁺ T cell division was visualized by monitoring 5,6-carboxy-fluorescein diacetate succinimidyl ester (CFSE) dilution after 48 h of co-culture. The division frequency of MataHari CD8 T cells in the presence of expressed HBx was similar to that of control CD8 T cells. Addition of IL-2 into the co-culture increased the number of dividing CD8 T cells regardless of HBx expression (Fig. 3). These results imply that HBx expression in hepatoctyes may not influence CD8⁺ T cell proliferation driven by hepatocyte surface antigen.

IFN-γ is involved in suppression of HBV transcription level and activation of immune cells important for antiviral immunity such as T cells and NK cells. We analyzed whether IFN-γ production was influenced by HBx expression of hepatocytes in co-culture with CD8 T cells. Compared to IFN-γ production in co-cultures of CD8 T cells with normal hepatocytes, IFN-γ production was increased in co-cultures of hepatocytes infected with baculoviruses, regardless whether HBx was expressed or not. However, IFN-γ production in co-cultures where HBx was expressed was lower than that in co-cultures with hepatocytes infected with control baculoviruses. IFN-γ production was barely detected in cultures of lymphocytes alone or hepatocytes alone (Fig. 4). These results imply that IFN-γ production may be suppressed when CD8 T cells are interacted with antigen on hepatocytes expressing HBx. IL-10 production was assessed in this co-culture system; however, it was always below detectable limits (data not shown).

MHC and ICAM-1 expression are elevated in human hepatoma cell lines through transfection of either the HBV or HBx genes (19-22). We reasoned that upregulation of these genes by HBx may not have occurred in our system based on the results of CD8 T cell proliferation and IFN-γ production. Appropriately, we evaluated transcriptional levels of H-2K and ICAM-1 in hepatocytes infected with baculoviruses that did and did not express HBx. As expected, enhanced H-2K transcription in hepatocytes was observed solely upon baculovirus infection; the expression of HBx did not further enhance H-2K transcription. Similarly, a transient increase of ICAM-1 transcription was detected in hepatocytes infected with baculoviruses regardless of whether HBx was expressed (Fig. 5). Finally, to clarify the nature of the enhanced apoptosis of CD8 T cells by HBx expression in hepatocytes, we examined hepatocytic transcription of PD-L1, a molecule implicated in apoptosis of CD8 T cells in the liver (3,15). Like H-2K and ICAM-1, PD-L1 transcription was upregulated by baculovirus infection but not further upregulated by HBx expression (Fig. 6A). To exclude the possibility that baculovirus infection may have masked the effect of HBx on PD-L1 induction, we compared PD-L1 mRNA levels in hepatocytes from HBx transgenic mice with those from normal mice. No differences were evident (Fig. 6B). These observations indicate that regulation of PD-L1 expression may not account for the increase of CD8 T cell apoptosis by HBx expression in hepatocytes.