Human Hepatitis B Viral e Antigen Interacts with Cellular Interleukin-1 Receptor Accessory Protein and Triggers Interleukin-1 Response*

Human hepatitis B virus (HBV) can cause acute and chronic hepatitis, cirrhosis, and hepatocellular carcinoma. HBV e antigen (HBeAg), a secreted protein and not required for viral replication, is thought to play an immunoregulatory role during viral infection. However, the functional involvement of HBeAg in host immune response has not been fully elucidated. We report in this study that HBeAg can bind to interleukin-1 receptor accessory protein (IL-1RAcP). Interleukin-1 (IL-1) plays an important role in inflammation and regulation of immune response, and membrane form of IL-1RAcP (mIL-1RAcP) is an essential component of trimeric IL-1/IL-1 receptor/mIL-1RAcP complex. We show that glutathione S-transferase- or polyhistidine-tagged recombinant HBeAg can interact with endogenous mIL-1RAcP in vitro. Purified (His)6-HBeAg added in the culture medium can interact with overexpressed FLAG-tagged mIL-1RAcP in vivo. Indirect immunofluorescence and confocal microscopy show that HBeAg colocalizes with mIL-1RAcP on the cell surface. Furthermore, HBeAg is able to induce the interaction of IL-1 receptor I (IL-1RI) with mIL-1RAcP and trigger the recruitment of adaptor protein myeloid differentiation factor 88 (MyD88) to the IL-1RI/mIL-1RAcP complex. Assembly and activation of IL-1RI/mIL-1RAcP signaling complex by HBeAg can activate downstream NF-κB pathway through IκB degradation, induce NF-κB-dependent luciferase expression, and induce the expression of IL-1-responsive genes. Silencing of IL-1RAcP by small interfering RNA dramatically abolishes HBeAg-mediated NF-κB activation. These results demonstrate that HBeAg can trigger host IL-1 response by binding to mIL-1RAcP. The interaction of HBeAg with mIL-1RAcP may play an important role in modulating host immune response in acute and chronic HBV infection.

Patients with persistent HBV infection are at a high risk of developing chronic hepatitis, cirrhosis, and hepatocellular carcinoma (HCC) (1,2). HBV e antigen (HBeAg), a 17-kDa secreted protein, is not a structural component of virion nor is it required for viral assembly, replication, or infection. However, conservation of HBeAg in genomes of all hepadnaviruses suggests an important role of HBeAg in HBV infection. The presence of HBeAg in the sera of patients generally indicates ongoing HBV replication and host liver disease. Most individuals with anti-e antibodies tend to have lower levels of viremia and remission of liver disease (3). However, a significant number of HBeAg-negative carriers have elevated levels of HBV and active liver disease. In these carriers, HBV genome usually harbors mutations that prevent the production of HBeAg (4 -6). These HBeAg-negative variants are often found in acute fulminant rather than chronic HBV infection in neonatal infection and in fulminant hepatitis than in self-limited hepatitis in adult acute infection (7)(8)(9)(10)(11). During chronic infection in some patients, the emergence of HBeAg-negative variants correlates with an exacerbation of liver injury and sometimes even with viral clearance (12)(13)(14)(15). In murine models, HBeAg preferentially elicits Th2-like response and has the potential to preferentially deplete HBeAg-and core-specific Th1 cells (16,17). These results suggest that HBeAg plays an immunoregulatory role and promotes viral persistence. With regard to hepatocellular carcinogenesis, HBeAg-positive patients have significantly elevated risk of developing HCC (18). Taken together, these findings suggest that HBeAg plays an important role in the chronicity and carcinogenesis of HBV infection. Despite these important clinical implications, the functional involvement of HBeAg in HBV infection and the molecular mechanism of HBeAg-host interactions remain largely unknown. In this study, we show that HBeAg can interact with host interleukin-1 receptor accessary protein (IL-1RAcP) and activate IL-1 signaling pathway.
The proinflammatory cytokine interleukin-1 (IL-1) initiates a wide spectrum of immunological responses to infection, stress, and tissue damage (19). IL-1 also functions as a costimulator to activate Th2 cells (20,21). Binding of IL-1 to type I IL-1 receptor (IL-1RI) leads to recruitment of membrane form of IL-1RAcP (mIL-1RAcP), which is essential for signal transduction (22). This is followed by recruitment of several intracellular adaptor proteins and kinases, including Toll-interacting protein (Tollip), MyD88, and members of the interleukin-1 receptor-associated kinase family. Tumor necrosis factor (TNF) receptor-associated factor 6 (TRAF6) is also recruited transiently to the receptor complex and triggers signal events that culminate in the activation of IB kinase complex and specific mitogen-activated protein kinase kinase kinases. Activated IB kinases phosphorylate the NF-B inhibitor IB, leading to the degradation of IB and activation of NF-B, whereas activated mitogen-activated protein kinase kinase kinases phosphorylate and activate c-Jun N-terminal kinase and p38 mitogen-activated protein kinase. IL-1 signal transduction ultimately leads to the activation of IL-1-responsive genes (19,23).
In this study, we show that HBeAg interacts and colocalizes with mIL-1RAcP on the cell surface. Furthermore, HBeAg is able to induce the interaction of IL-1RI with mIL-1RAcP, trigger the recruitment of adaptor protein MyD88 to the IL-1RI/mIL-1RAcP complex, activate NF-B through IB-␣ degradation, induce NF-B-dependent reporter gene expression, and induce the expression of IL-1-responsive genes. Silencing IL-1RAcP dramatically abolishes HBeAgmediated NF-B activation. Our observations reveal an important function of HBeAg. The physiological significance of this interaction is discussed.
Expression Vectors-For bacterial expression, the full-length HBeAg was fused downstream of a polyhistidine tag of a pRSETB vector. The full-length HBeAg and different segments of core protein, respectively, were fused downstream of a GST tag of a pGEX-3X vector. For in vitro transcription and translation reaction, the #46 cDNA was inserted downstream of an influenza viral HA epitope in a pGEM-HA vector. For cytoplasmic expression in mammalian cells, the full-length HBeAg and different segments of core protein, respectively, were fused downstream of a GST tag of a pCMV-GST vector. The #46 protein, sIL-1RAcP356, and MyD88, respectively, were fused downstream of an HA epitope of a pCMV-HA vector. For membrane expression, the mIL-1RAcP and IL-1RI, respectively, were fused downstream of the FLAG tag of a pFLAGCMV1 vector or the HA tag of a pHACMV1 vector.
Yeast Two-hybrid Screen-The HBV core protein (adw subtype) from amino acids 119 -185, designated as 2/5C (Fig. 1A), was fused downstream of the LexA DNA-binding domain of a pLexA-BD vector and used as a bait to screen a human liver cDNA library. A total of 5 ϫ 10 7 cfu was screened.
Preparation of Recombinant Proteins and Bead-bound Proteins-Expression of the 19-kDa (His) 6 -HBeAg was induced by 1 mM isopropyl 1-thio-␤-D-galactopyranoside in bacterial strain Escherichia coli BL21(DE3). Recombinant (His) 6 -HBeAg protein was purified under native condition on a nickel-agarose column and an anion exchanger column. The purification process was performed by fast protein liquid chromatography. The protein was Ͼ95% pure as judged by SDS-PAGE followed by silver staining. Fifty g of purified (His) 6 -HBeAg in 500 l of binding buffer (5 mM imidazole, 0.5 M NaCl, 20 mM Tris-HCl (pH 7.9)) was incubated with 40 l of nickel-charged His.Bind beads. The beads were washed three times with washing buffer (60 mM imidazole, 0.5 M NaCl, 20 mM Tris-HCl (pH 7.9)). Recombinant GST fusion proteins, GST-2/5C, GST-Cd2/5C, and GST-HBeAg, were induced by 1 mM isopropyl 1-thio-␤-Dgalactopyranoside in bacterial strain RRI. Approximately 0.3 mg of GST fusion proteins in 1 ml of phosphate-buffered saline (PBS) was incubated with 40 l of glutathione-conjugated agarose beads (50% slurry). The beads were washed three times with PBS containing 0.5% Triton X-100.
In Vitro Interaction-In vitro transcription and translation reactions were performed with TNT quick-coupled transcription/translation system in the presence of [ 35 S]methionine according to the manufacturer's recommendation. In vitro transcription/translation products were incubated with glutathione bead-bound GST, GST-2/5C, and nickel bead-bound (His) 6 -HBeAg, respectively, in binding buffer (25 mM HEPES (pH 7.9), 12.5 mM Mg 2 Cl, 100 mM KCl, 0.1 mM EDTA, 0.5 mM dithiothreitol, and 0.1% Nonidet P-40). The beads were then washed four times in 0.5% Triton X-100 in PBS. Bead-bound proteins were eluted with SDS-PAGE sample buffer, resolved on a 12% SDS-PAGE, and detected by autoradiography.
Cell Culture and Transfection-Human embryonic kidney 293T (HEK293T) cells were grown in Dulbecco's modified eagle's medium supplemented with 10% bovine calf serum, 10 2 units/ml penicillin, and 10 g/ml streptomycin at 37°C at 8% CO 2 . Human hepatoma HA22T/VGH cells were grown in Dulbecco's modified eagle's medium supplemented with 10% fetal bovine serum, 2 mM glutamine, 0.1 mM non-essential amino acid, 10 2 units/ml penicillin, and 10 g/ml streptomycin at 37°C at 5% CO 2 . Plasmid DNAs were transfected into HEK293T cells or HA22T/VGH cells in a 10-cm dish by the calcium phosphate method. Each set of experiments was performed with two different preparations of plasmids and repeated two to three times for each preparation.
Immunoprecipitation-Cells in a 10-cm dish were lysed in 400 l of 0.1% Nonidet P-40 lysis buffer (0.1% Nonidet P-40, 250 mM NaCl, 5 mM dithiothreitol, 1 mM EDTA, 10% glycerol) containing Complete TM protein inhibitor mixture on ice for 2-3 h. Twenty l of M2 affinity beads was added to the whole-cell extracts and incubated for 3 h at 4°C with rocking. Alternatively, 40 l of anti-core antibody was added to the whole-cell extracts and incubated overnight at 4°C with rocking. After adding 20 l of protein A-agarose slurry, the mixtures were incubated for 3 h at 4°C with rocking. The M2 beads and anti-core protein A beads, respectively, were washed three times with 0.1% Nonidet P-40 lysis buffer.
Western Blot-Proteins were transferred to Hybond ECL or Hybond-P filter and detected by using primary antibody and secondary antibody. The immunoreactive proteins were visualized with Western Lightning Chemiluminescence Reagent Plus.
Indirect Immunofluorescence and Confocal Microscopy-Transfected HEK293T cells were precooled for 20 min at 4°C and then incubated with HBeAg or IL-1␤ for 1 h at 4°C. After washing three times with PBS, cells were incubated with PBS containing 1% paraformaldehyde and 1% sucrose for 10 min at room temperature. Fixed cells were washed twice with PBS, blocked by incubation with 10% bovine serum albumin (BSA) in PBS for 30 min at 37°C, and incubated with primary antibody in 3% BSA/PBS for 1 h at 37°C. After extensive washing with PBS, cells were incubated with anti-mouse IgG-TRITC conjugated and anti-rabbit IgG-FITC conjugated secondary antibodies and H33342 in 3% BSA/PBS for 45 min at 37°C. After extensive washing with PBS, cells were examined by laser confocal microscopy.
Electrophoretic Mobility Shift Assay-Nuclear extracts of HA22T/VGH cells were prepared as described previously (25). Electrophoretic mobility shift assays were performed by incubating 10 g of nuclear extracts with 1 ϫ 10 5 cpm of [␥-32 P]ATP end-labeled double-stranded wild type NF-B oligonucleotides (5Ј-AGTTGAGGGGACTTTCCCAGG-3Ј) in 20 l of binding buffer at 30°C for 30 min as described previously (25). For competition experiments, unlabeled wild type or mutant NF-B oligonucleotides (5Ј-AGTTGAGGCGACTT-FIGURE 1. A, structure of HBeAg, core protein, and various core deletion mutants. HBeAg and core share a large stretch of sequence in common and only differ at the N and C termini. Deletion mutants 2/5C, 1/5C, 2/5Cd1/5C, and Cd2/5C contain the sequence of core protein from amino acids 119 -185, 144 -185, 116 -149, and 1-117, respectively. B, structure of the membrane and soluble forms of human IL-1RAcP. Both soluble forms, sIL-1RAcP356 and sIL-1RAcP␤, contain the extracellular domain but lack the intracellular and transmembrane domains of the membrane form, mIL-1RAcP. The protein encoded by #46 cDNA corresponds to amino acids 247-356 of sIL-1RAcP356. NOVEMBER 10, 2006 • VOLUME 281 • NUMBER 45 ]methionine-labeled HA or HA-46 was incubated with glutathione bead-bound GST and GST-2/5C and nickel bead-bound (His) 6 -HBeAg, respectively. After extensive wash, bead-bound proteins were analyzed by autoradiography. C, in vitro interaction of HBeAg with endogenous mIL-1RAcP. Pulldown experiments of precleared lysates of HEK293T cells were performed with glutathione bead-bound GST-2/5C and GST-HBeAg and nickel bead-bound (His) 6 -HBeAg, respectively. The endogenous mIL-1RAcP in pulldown products was detected by anti-IL-1RAcP antibody. The bead-bound GST-HBeAg and (His) 6 -HBeAg were detected with anti-core antibody. Bead-bound GST-2/5C could not be detected by anti-core antibody as the relevant epitope was lacking. D, interaction of exogenous (His) 6 -HBeAg added into culture medium with the overexpressed mIL-1RAcP. HEK293T cells were transfected with FLAG or f:mIL-1RAcP. The purified (His) 6 -HBeAg was added into the culture medium at 4°C for 20 min. HBeAg in lysates was immunoprecipitated down by anti-core protein A beads. The f:mIL-1RAcP protein in the immunoprecipitates and in whole-cell extracts was detected by anti-FLAG antibody. (His) 6 -HBeAg in the immunoprecipitates was detected by anti-core antibody. (His) 6 -Im, a His-tagged bacterial immunity protein of irrelevant function, served as a negative control. E, colocalization of HBeAg with mIL-1RAcP on cell surface. HEK293T cells were transiently transfected with f:mIL-1RAcP or f:IL-1RI. After incubation with HBeAg or IL-1␤ for 1 h at 4°C, cells were fixed with 1% paraformaldehyde for 10 min at room temperature. After blocking, cells were incubated with primary antibody for 1 h at 37°C followed by anti-mouse IgG-TRITC-conjugated and anti-rabbit IgG-FITC-conjugated secondary antibodies and H33342 for 45 min at 37°C and analyzed by laser confocal microscopy.

HBeAg and IL-1RAcP
TCCCAGGC-3Ј) were added to the binding reaction on ice 5 min before addition of the radiolabeled probe.
Luciferase Reporter Assay-HA22T/VGH cells were cotransfected with plasmids containing NF-B-driven firefly luciferase and TK-driven Renilla luciferase. Twenty-four h posttransfection, cells were treated with different reagents. To assay firefly and Renilla luciferase activities, cells were lysed using passive lysis buffer, and luciferase activities were determined by the manufacturer's standard protocols. Firefly luciferase activity values were normalized to Renilla luciferase activity values to control for transfection efficiency. Luciferase activity was measured in duplicate wells, and experiments were repeated at least four times.
ELISA-HA22T/VGH cells were seeded in 24-well flat-bottom plates at 2 ϫ 10 6 cells/ml and cultured for 2 days before stimulation. Cells were washed three times with PBS and stimulated with different reagents. Culture media were collected at 15 h for TNF-␣ and 24 h for GM-CSF, respectively, after stim-ulation. Levels of TNF-␣ and GM-CSF in the culture media were measured by ELISA. ELISA was performed by adding 50 l of each sample to a 96-well plate of TNF-␣ and GM-CSF ELISA kits according to the manufacturer's recommendation. Each experiment was performed in triplicate wells and final results were the average of three independent experiments.

RESULTS
In Vivo and in Vitro Interaction between HBeAg and IL-1RAcP-HBeAg shares a large stretch of sequence in common with HBV core protein, which assembles to form viral capsid. They only differ at the N and C termini (Fig. 1A). To identify the cellular protein(s) that interact with HBeAg and/or core protein, the C-terminal 67 amino acid-region from amino acid 119 to 185 of core protein (2/5C) was used as a bait to screen the human liver cDNA library in a yeast two-hybrid system. As shown in Fig. 1B, a ϳ0.6-kb #46 cDNA was identified and confirmed to interact with the bait in yeast two-hybrid system. Sequence analysis of #46 shows it corresponds to amino acids 247-356 segment of the soluble form of human IL-1RAcP, sIL-1RAcP356. For human IL-1RAcP, one 570amino acid membrane form (mIL-1RAcP) and two soluble forms, 356-amino acid form (sIL-1RAcP356) and 346-amino acid form (sIL-1RAcP␤) that are generated by alternative splicing have been identified (22,26,27). Both soluble forms lack the intracellular and transmembrane domains of the membrane form. The sequence of mIL-1RAcP and sIL-1RAcP356 is identical from amino acid 1 to 350. The N-terminal segment of HBeAg and IL-1RAcP NOVEMBER  sIL-1RAcP␤, from amino acid 1 to 301, is identical to that of sIL-1RAcP356. A unique second half of the Ig3 domain makes up the rest of sIL-1RAcP␤. mIL-1RAcP is essential for IL-1mediated response such as activation of interleukin-1 receptorassociated kinase, NF-B, c-Jun N-terminal kinase, and p38 mitogen-activated protein kinase (19,22,23). sIL-1RAcP356 has been shown to act as an inhibitor of IL-1 signaling (26,28). The function of sIL-1RAcP␤ is still unknown (27).
Binding of HBeAg to mIL-1RAcP on the Cell Surface-HBeAg is a secreted protein produced during HBV replication, and present in the sera of the patients in natural infection. The above results raise the intriguing possibility that HBeAg in the sera of the patients may interact with the extracellular domain of a membrane-bound IL-1RAcP protein. To determine whether there is an interaction between endogenous mIL-1RAcP with HBeAg in vitro, crude HEK293T lysates were precleared with glutathione bead-bound GST-Cd2/5C protein to remove proteins that could bind to GST or the portion of core protein not implicated in the interaction with IL-1RAcP. The precleared lysates were then incubated with glutathione beadbound GST, GST-2/5C and GST-HBeAg, and nickel beadbound (His) 6 -HBeAg, respectively. The endogenous mIL-1RAcP was pulled down by GST-2/5C, GST-HBeAg, and (His) 6 -HBeAg but not by GST (Fig. 2C). To further confirm whether the HBeAg can interact with mIL-1RAcP in vivo, purified (His) 6 -HBeAg was added into the culture medium of HEK293T cells, which was transiently transfected to overexpress the membrane FLAG:mIL-1RAcP protein, for 20 min. (His) 6 -HBeAg in lysates was immunoprecipitated with anticore antibody and protein A-agarose beads. FLAG:mIL-1RAcP was coimmunoprecipitated with (His) 6 -HBeAg, suggesting the interaction between exogenous (His) 6 -HBeAg added to the culture medium and mIL-1RAcP on the cell surface (Fig. 2D). To further demonstrate that the binding of HBeAg to mIL-1RAcP indeed takes place on the cell surface, indirect immunofluorescence followed by confocal microscopy was performed on HEK293T cells transiently transfected to overexpress the membrane FLAG:mIL-1RAcP protein. As shown in Fig. 2E, binding of fluorescence labeled HBeAg to the cell surface of HEK293T cells was observed. The decoration of cell surface by HBeAg was only noted in HEK293T cells that also overexpressed FLAG: mIL-1RAcP. HBeAg binding intensity paralleled the level of expression of FLAG:mIL-1RAcP in the transfectants. Moreover, colocalization of HBeAg and FLAG:mIL-1RAcP in transfected HEK293T cells was clearly demonstrated by confocal microscopy. Binding of IL-1␤ to HEK293T cells overexpressing the membrane FLAG:IL-1RI protein displayed the same pat- HA22T/VGH cells were transfected with siRNA targeting the N-terminal region of IL-1RAcP or non-targeting control siRNA for 72 h followed by treatment with 3 g/ml HBeAg, 10 ng/ml IL-1␤, or 1 g/ml LPS for 1 h. Expression of endogenous mIL-1RAcP or ␤-actin mRNA was determined by RT-PCR, while NF-B activation was determined by electrophoretic mobility shift assay. D, enhancement of IL-1␤-mediated NF-B activation by HBeAg. HA22T/VGH cells were treated with 1 g/ml HBeAg, 3 ng/ml IL-1␤, or in combination for 1 h. Pretreatment was performed with 10 g/ml polymyxin B for 30 min. tern, which served as a positive control. These observation establishes specific interaction between HBeAg and FLAG: mIL-1RAcP on the cell surface.
Association of IL-1RI and mIL-1RAcP and Recruitment of Adaptor Protein MyD88 Induced by HBeAg Binding-The demonstration that HBeAg binds to mIL-1RAcP on the cell surface raises the intriguing possibility that HBeAg can trigger the IL-1-mediated signal transduction. We first tested whether HBeAg could induce the association of IL-1RI and mIL-1RAcP. Human hepatoma HA22T/VGH cells overexpressing the membrane HA:IL-1RI and FLAG:mIL-1RAcP proteins were treated with HBeAg. These cells were also stimulated with IL-1␤ as a positive control. As shown in Fig. 3A, the association of IL-1RI and mIL-1RAcP was strongly induced by HBeAg, the association of which was also induced by IL-1␤. To ensure that the association between IL-1R1 and mIL-1RAcP was triggered by HBeAg itself rather than contaminants from the HBeAg preparation, we did the following two control experiments: 1) this association was not affected by pretreatment with polymyxin B, a LPS-specific binding antibiotic. This demonstrates that this effect does not result from LPS, which is a possible contaminant carried with the HBeAg preparation; 2) this association could be abolished by boiling of HBeAg. Boiling can denature HBeAg but not LPS. Together, these results demonstrate that HBeAg is able to induce the association of IL-1RI and mIL-1RAcP. Adaptor protein MyD88 is recruited into the IL-1RI and mIL-1RAcP complex when association of IL-1RI and mIL-1RAcP is induced by IL-1␤. To test whether MyD88 is recruited into the IL-1RI and mIL-1RAcP complex by treatment with HBeAg, HA22T/VGH cells overexpressing FLAG: IL-1RI, FLAG:mIL-1RAcP, and HA:MyD88 were treated with either HBeAg or IL-1␤ with the latter serving as a positive control. As shown in Fig. 3B, recruitment of MyD88 to the IL-1RI and mIL-1RAcP complex could be induced by HBeAg. This observation was not affected by pretreatment with polymyxin B, demonstrating that HBeAg is able to induce the recruitment of MyD88 to the IL-1RI and mIL-1RAcP complex. These results strongly indicate that HBeAg can trigger the IL-1 signaling pathway.
Induction of IB-␣ Degradation and Increase of NF-B DNA Binding by HBeAg-IL-1␤ induces the phosphorylation and degradation of IB leading to release and activation of NF-B. HBeAg-induced NF-B activation was assessed by electrophoretic mobility shift assay. As shown in Fig. 4, A and B, HBeAg increased the NF-B DNA binding activity, which was specifically competed by wild type NF-B but not mutant NF-B oligonucleotides. This correlated with an increased degradation of IB-␣. NF-B DNA binding activity and IB-␣ degradation induced by HBeAg was not affected by polymyxin B pretreatment but abolished by boiling of HBeAg. In contrast, the LPS-induced IB-␣ degradation and NF-B DNA binding activity were abolished by polymyxin B pretreatment but not affected by boiling. These results demonstrate that HBeAg can induce degradation of IB-␣ and NF-B activation.
Abolishment of HBeAg-induced NF-B Activation by Silencing of Endogenous IL-1RAcP-To further determine the role of mIL-1RAcP on HBeAg-induced NF-B activation, specific siRNA was designed and used to silence the expression of endogenous IL-1RAcP in HA22T/VGH cells. As shown in Fig.  4C, expression of endogenous mIL-1RAcP mRNA was dramatically reduced by IL-1RAcP specific siRNA, compared with control siRNA. When endogenous mIL-1RAcP mRNA expression was greatly diminished, the effect of HBeAg on NF-B activation was almost completely abolished. These results demonstrate that endogenous mIL-1RAcP plays an important role in HBeAg-induced NF-B activation. The IL-1␤-induced NF-B activation was also abolished by silencing of mIL-1RAcP mRNA expression, which was consistent with the essential role of mIL-1RAcP in IL-1-mediated response. In contrast, the LPSinduced NF-B activation was not affected by silencing of mIL-1RAcP mRNA expression, which was consistent with the previous finding that mIL-1RAcP plays no role in LPS-induced NF-B activation.
Enhancement of IL-1␤-mediated NF-B Activation by HBeAg-The signaling cascade initiated by IL-1␤ binding to IL-1RI involves the assembly of a trimeric protein complex consisting of IL-1, IL-1RI, and mIL-1RAcP. Our results shown above demonstrate that HBeAg binds to mIL-1RAcP and activates the IL-1 signal transduction pathway. It is therefore important to know whether HBeAg can cooperate with IL-1␤ in the induction of IL-1 signal transduction pathway. As shown in Fig. 4D, higher NF-B DNA binding activity was noted with the combination of HBeAg and IL-1␤ than either stimulus alone. This was not abolished by polymyxin B pretreatment. The results demonstrate that HBeAg enhances IL-1␤-mediated NF-B activation.
Expression of NF-B-dependent Reporter Gene Induced by HBeAg-We then examined whether NF-B-dependent reporter gene expression could be induced by HBeAg. As shown in Fig. 5, HBeAg treatment increased the levels of the NF-B-dependent luciferase activity values by 2.40-fold, compared with that of 2.39-fold induced by IL-1␤. This HBeAg effect was not affected by polymyxin B pretreatment but abolished by boiling of HBeAg.
Expression of IL-1␤-responsive Genes Induced by HBeAg-IL-1 can induce the expression of multiple genes, including IL-1␤, IL-6, TNF-␣, iNOS, MIP-1␣, and GM-CSF. We examined whether the expression of these genes in HA22T/VGH cells was induced by HBeAg by RT-PCR first. As shown in Fig.  6, HBeAg induced the expression of all of these IL-1-responsive genes. This HBeAg effect was not affected by polymyxin B pretreatment but abolished by boiling of HBeAg. In contrast, LPS only induced the expression of IL-1␤, IL-6, TNF-␣, and GM-CSF. The LPS effect was abolished by polymyxin B pretreat-

DISCUSSION
In this study, we demonstrate that HBeAg can bind to mIL-1RAcP on the cell surface, induce the association of IL-1RI and mIL-1RAcP, trigger the recruitment of adaptor protein MyD88 to the IL-1RI and mIL-1RAcP complex, activate NF-B through the IB-␣ degradation, and induce the expression of IL-1-responsive genes including IL-1␤, IL-6, TNF-␣, iNOS, MIP-1␣, and GM-CSF. Since the serum concentration of HBeAg in chronic HBV carriers is from 2.3 to 16 g/ml (29), our observations using the concentration of HBeAg from 1 to 24 g/ml is expected to be of physiological relevance in vivo. Our results strongly suggest that HBeAg may induce the IL-1 signaling and modulate host immune response through its interaction with mIL-1RAcP during natural infection of human hepatitis B virus.
In HBV infection, HBeAg-negative variants are not only associated with acute fulminant hepatitis in both neonatal and adult infection but also correlates with acute exacerbation of liver inflammation and viral clearance in chronic infection. These clinical observations indicate that HBeAg may attenuate the cell-mediated inflammatory response (12)(13)(14)(15). In animal studies, HBeAg has also been shown to preferentially elicit Th2like response and deplete HBeAg-and core-specific Th1 cells in mice (16,17). HBeAg, therefore, has been proposed to play an immunoregulatory role and promote viral persistence. However, the molecular mechanism of the immunoregulatory function of HBeAg is still largely unknown. We show here that HBeAg can bind to mIL-1RAcP and induce the IL-1 signaling. IL-1 has been shown to function as a costimulator to activate Th2 cells (20,21). Other IL-1-responsive proteins such as IL-6 also exert anti-inflammatory effect and/or promote Th2 differentiation (30). Our results therefore indicate that HBeAg secreted from the HBVinfected hepatocytes during HBV replication may tip the immune response to a Th2-like response, leading to suppression of the host immune response against HBV and thus preventing viral clearance and promoting viral persistence.
HBeAg positivity, an indicator of active HBV replication, is associated with an increased risk of HCC (18). In the past, the role of HBeAg in the development of HCC has been explained by its indication of active HBV replication. Active replication of HBV may directly initiate malignant transformation by increasing the expression of X protein of HBV or by increasing the probability of insertion of viral DNA in or near tumorsuppressor genes or proto-oncogenes (31,32). Indirectly, active replication of HBV causes chronic necroinflammatory disease, which accelerates the hepatocyte turnover and generates mutagenic reactive oxygen species during the inflammatory process. The latter leads to the accumulation of DNA damage or insertion of viral DNA into chromosomal DNA. Our finding that HBeAg can activate IL-1 signaling shows that HBeAg may itself contribute to the development of HCC. First, activation of NF-B-dependent proinflammatory cytokines may enhance the inflammatory process, which induces the hepatocyte proliferation (33). Second, NF-B activation can induce anti-apoptotic genes, which may protect the transformed hepatocytes against apoptosis and promote the carcinogenesis of hepatocytes (34). The essential role of NF-B in promoting the development of HCC from inflammatory hepatitis in a mouse model has recently been demonstrated (35). Moreover, constitutive activation of NF-B has often been observed in cancer tissues of HBV-positive HCC (36). In addition, recent evidence also shows that IL-1 prevents apoptosis in many cell types including keratinocytes, chondrocytes, osteoclasts, neutrophils, monocytes, and lymphocytes (37)(38)(39)(40)(41).
IL-1 has a broad range of biological effects on different types of cell in response to infection, tissue damage, and stress. It also acts as a costimulator to activate Th2 cells (20,21). Different viral proteins have been shown to regulate the IL-1 response through different mechanisms. The CrmA protein of cowpox virus inhibits the production of caspase-1, which prevents the proteolytic cleavage of pro-interleukin-1␤ (pro-IL-1␤) to mature IL-1␤ (50). B15R of vaccinia virus is a soluble IL-1␤ receptor and inactivates the IL-1 (51). The A46R and A52R proteins of vaccinia virus have been shown to inhibit IL-1 response by possessing putative TIR domain to suppress TIR domain-dependent signaling (52). N1L protein of vaccinia virus inhibits IL-1 response by targeting several components of the multisubunit IB kinase complex, especially the TANK-binding kinase 1 (TBK1) (53). Results of this study show that HBeAg of HBV uses a novel mechanism to modulate the IL-1 response by interacting with the mIL-1RAcP component of IL-1RI/mIL-1RAcP complex to induce the IL-1 response.
HBV appears to employ several strategies to promote its persistence during infection. Viral X protein has been shown to inhibit cellular proteasome activity and may inhibit antigen processing and presentation (54). During HBV replication, 22-nm surface antigen particles, which are secreted in a large amount from infected hepatocytes into blood stream, can consume neutralizing antibody against HBV. These surface antigen particles, in addition, may also function as a high dose tolerogen to suppress immune elimination of infected hepatocytes (55,56). Here we describe that HBeAg interacts with mIL-1RAcP to have a novel immunoregulatory function. The immunomodulation by HBeAg may represent a major function of HBeAg in the HBV life cycle and valuable target for potential therapeutic intervention.