Hepatitis B Virus Nucleocapsid but Not Free Core Antigen Controls Viral Clearance in Mice

Abstract

We have recently shown that hepatitis B virus (HBV) core antigen (HBcAg) is the major viral factor for HBV clearance using a hydrodynamics-based mouse model. Knockout of HBcAg hampers the development of antiviral immune responses and thus promotes HBV persistence. Here, we further demonstrated that only in the capsid form, but not the free or dimer form, can HBcAg exert its contributory role in HBV clearance. HBcAg is the main structural protein of HBV icosahedral nucleocapsid. A mutant HBV DNA which expresses an assembly-defective HBcAg, HBcAgY132A, surprisingly prolonged HBV surface antigenemia in both C57BL/6 and BALB/c mice without affecting viral transcription and translation. This result was not due to a loss of the possible immune epitope caused by the single-amino-acid substitution of HBcAg. Moreover, the particular HBV mutant failed to induce robust humoral and cellular immunity against HBV. These data revealed the requirement of capsid structure for inducing adequate immunity that leads to HBV clearance in mice.

INTRODUCTION

Chronic hepatitis B virus (HBV) infection has been known to be a major cause of hepatocellular carcinoma (HCC). Although many efforts have been made to unravel both viral and host factors involved in viral clearance or persistence, still little is known about the molecular and immune mechanisms of how HBV differentially leads to chronic infection or clearance.

We previously explored the HBV genes contributing to its persistence or clearance using a hydrodynamics-based mouse model and identified HBV core antigen (HBcAg) as the most critical viral factor for HBV clearance (18). The absence of HBcAg hampered the development of HBV-specific antiviral immune responses and significantly promoted HBV persistence in mice. HBcAg, the capsid protein of HBV, assembles into the icosahedral capsid particles in the T = 3 and T = 4 arrangement by 90 or 120 homodimers, respectively. During the assembly process, HBV pregenomic RNA (pgRNA) along with the viral polymerase (pol) is specifically incorporated into the virus particle to form nucleocapsids. The encapsidated pgRNA is reverse transcribed to DNA by the viral pol and then completes the synthesis of viral DNA. Therefore, HBcAg not only serves as the major structural protein of HBV but also participates in viral replication. Nevertheless, from the examination of a series of HBV mutants in the mouse model, we excluded the association between viral replication and HBV clearance because the injected HBV DNA persisted in the liver of mice regardless of viral replication. Anti-HBc antibodies did not play a major role in the determination of HBV clearance, either. How HBcAg functions and interacts with the host to elicit adequate antiviral immunity still remains unclear.

Recently, the viral capsid has been coincidentally demonstrated to function as a pathogen-associated molecular pattern (PAMP) of adenovirus (6) and retrovirus (20, 22) to trigger host innate immune signaling. Besides, TRIM5 (22) and cyclophilin A (20) were suggested as the intracellular pattern recognition receptors (PRRs) for the retroviral capsids but not free capsid proteins (25, 28). Given that HBcAg contributes to HBV clearance in vivo, it is worth further clarifying whether HBcAg has to be in the capsid form to contribute to viral clearance. In this study, we used an assembly-deficient HBV mutant, HBcY132A pAAV/HBV1.2, to investigate whether HBV capsid structure is important for viral clearance in mice. The mutant was first characterized in vitro to confirm the expression of a full-length but assembly-defective HBcAgY132A (5). Interestingly, this HBcY132A mutant resulted in a prolonged HBV persistence in mice without affecting the expression of other viral genes. Furthermore, impaired HBV-specific immune responses were observed in these mice. Our results suggested that the capsid structure of HBcAg is required for HBcAg to contribute to viral clearance.

MATERIALS AND METHODS

Plasmids.

To generate HBcY132A pAAV/HBV1.2 and pFLAG-CMV2/HBcY132A, in situ site-directed mutagenesis was performed using the QuikChange II site-directed mutagenesis kit (Stratagene). The paired primers used for the mutagenesis are as follows: HBcY132A-F, 5′TCGCACTCCTCCAGCCGCTAGACCACCAAATGC3′, and HBcY132A-R, 5′GCATTTGGTGGTCTAGCGGCTGGAGGAGTGCGA3′ (mutation sites shown in bold). The pAAV/HBV1.2 and pFLAG-CMV2/HBc plasmids (18) were used as a template for the generation of HBcY132A pAAV/HBV1.2 and pFLAG-CMV2/HBcY132A, respectively. HBeAg/core-null pAAV/HBV1.2 (with a premature stop codon at the 38th amino acid of HBcAg) and HBc175 pAAV/HBV1.2 (with a premature stop codon at the 176th amino acid of HBcAg) were described in reference 18.

Cell culture and transfection.

HuH-7 cells were maintained in Dulbecco's modified Eagle's medium (Biological Industries) containing 10% fetal bovine serum (Biological Industries) at 37°C in a 5% CO₂ atmosphere. Transfection was done using Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. Two days after transfection, the supernatant was collected to determine the levels of HBsAg and HBeAg, and cells were lysed to extract RNA or total proteins for Northern blot or Western blot analysis, respectively. To detect HBV nucleocapsid or capsid-associated DNA, cytoplasmic lysates were prepared 4 days after transfection.

Mice and hydrodynamic injection.

Six- to 7-week-old C57BL/6 or BALB/c male mice from the breeding colonies of National Taiwan University were used for hydrodynamic injection as described previously (18). Briefly, 10 μg of plasmid DNA in phosphate-buffered saline (PBS) was intravenously injected into the anesthetized mice in a volume equivalent to 8% of the body weight within 5 to 6 s. DNA samples for hydrodynamic injection were all purified using the EndoFree Maxi plasmid kit (Qiagen). For the trans-complementation assay, 10 μg of HBeAg/core-null pAAV/HBV1.2 was coinjected with 10 μg of pFLAG-CMV2/HBc or pFLAG-CMV2/HBcY132A mutant as indicated. At the indicated time points, the blood samples and liver tissues were collected to determine the serological markers for HBV and to examine viral expression, respectively.

The mouse experiments were performed with the approval of the Institutional Animal Care and Use Committee at the National Taiwan University College of Medicine.

Southern and Northern blotting.

Fifty milligrams of mouse liver tissue was lysed in 700 μl of lysis buffer (10 mM Tris-HCl, pH 8.0, 100 mM NaCl, 1 mM EDTA, 0.5% SDS, 0.5 mg of proteinase K) at 37°C overnight. Twenty micrograms/milliliter of RNase A was added 1 h before extraction. Total DNA was extracted by phenol and chloroform and then dissolved in 0.1× TE buffer (1 mM Tris-HCl, pH 8.0, 0.1 mM EDTA) with 20 μg/ml of RNase A. The cytoplasmic core-associated DNA from cells prepared as described in reference 32 or 30 μg of total DNA from a mouse liver tissue was used for Southern blot analysis of HBV DNA. Mouse DNA was digested with 40 units of HindIII (NEB) at 37°C overnight before gel electrophoresis.

RNA was extracted by TRIzol (Invitrogen) according to the manufacturer's instructions. One milliliter of TRIzol was used for 50 mg of mouse liver tissue or cells per well of a 6-well cell culture plate. Ten micrograms of RNA from each sample was used for Northern blotting.

Southern and Northern blot analyses were performed using digoxigenin (DIG)-labeled probes as previously described (15). DIG-HBx, DIG–mouse glyceraldehyde-3-phosphate dehydrogenase (DIG-mGAPDH), and DIG-human GAPDH (DIG-hGAPDH) probes were used to detect HBV DNA/RNA and mouse and human GAPDH mRNA, respectively.

Western blotting.

Cells or 50 mg of liver tissue was lysed in RIPA buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% Triton X-100, 5 mM EDTA, 0.25% sodium deoxycholate, 0.1% SDS) containing the Complete protease inhibitor (Roche). Then, 50 or 100 μg of protein lysates was subjected to SDS-PAGE followed by immunoblotting using rabbit anti-HBcAg (1:1,000; generated by LTK BioLaboratories), mouse anti-β-actin (1:5,000; Sigma), mouse anti-Flag (1:1,000; Sigma), horseradish peroxidase (HRP)-conjugated goat anti-rabbit (1:5,000; Promega), or rabbit anti-mouse (1:5,000; Dako). Chemiluminescence detection was performed with Immobilon Western chemiluminescent HRP substrate (Millipore).

Detection of HBV nucleocapsid.

To detect intracellular HBV capsids, HuH-7 cells transfected with indicated DNAs were lysed using NET buffer (50 mM Tris-HCl, pH 8.0, 1 mM EDTA, 100 mM NaCl, 0.5% NP-40) 4 days after transfection. After incubation at 4°C for 1 h with agitation, cytoplasmic lysate was cleared by centrifugation at 12,000 rpm for 20 min at 4°C. To prepare the cytoplasmic lysate of mouse liver tissue, 50 mg of liver tissue was lysed by 700 μl of tissue lysis buffer (10 mM Tris-HCl, pH 8.0, 150 mM NaCl, 0.5% NP-40, Complete protease inhibitor [Roche]) on ice for 1 h and was centrifuged to remove the nuclei and cell debris. The lysate was further treated with 120 units of micrococcal nuclease (Fermentas) and 6 mM CaCl₂ at 37°C for 30 min, and 25 mM EDTA was added to terminate the reaction. Twenty-five micrograms of lysate was separated by electrophoresis in 1% agarose gel in 1× Tris-borate-EDTA (TBE) buffer (Amresco) and then was electrotransferred onto a polyvinylidene difluoride (PVDF) membrane (for Western blotting) or a positively charged nylon membrane (Roche) (for Southern blotting) in TNE buffer (10 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1 mM EDTA). To visualize the capsid particles, the PVDF membrane was subjected to Western blotting using anti-HBc (1:5,000). To detect capsid-associated viral DNA, the transferred nylon membrane was incubated with denaturing buffer (0.2 M NaOH, 1.5 M NaCl) for 5 min and neutralizing buffer (0.2 M Tris-HCl, 1.5 M NaCl, pH 7.5) for 5 min and was subjected to Southern blot analysis as described above.

Detection of serum or culture supernatant HBV antigens and antibodies.

The levels of HBsAg, HBeAg, anti-HBs, and anti-HBc antibodies in the culture supernatant or mouse sera were measured using the Axsym system kits (Abbott). According to the supplier's criteria, the positivity criteria of HBsAg, HBeAg, anti-HBc, and anti-HBs are signal/noise (S/N) ratio of ≥2, signal/cutoff (S/CO) ratio of ≥1, S/CO ratio of <1, and mIU/ml of ≥10, respectively.

Immunohistochemical staining.

At indicated time points, the liver samples were collected and embedded in optimal cutting temperature (OCT). A cryosection of each sample was prepared for staining of HBcAg and HBsAg using anti-HBc (1:1,000; Dako) or anti-HBs (1:800; Biomeda) antibodies and Envision_System, HRP (diaminobenzidine [DAB]) (Dako). Hematoxylin was used to visualize the nuclei.

IFN-γ ELISPOT.

At 10 days postinjection (dpi), the splenocytes of mice receiving the indicated DNA were isolated. After stimulation with 0.3 μg/ml of recombinant HBcAg (ID Labs), gamma interferon (IFN-γ)-secreting cells were detected using an IFN-γ enzyme-linked immunosorbent spot (ELISPOT) assay set (BD Biosciences) as described previously (18).

RESULTS

Characterization of HBcY132A pAAV/HBV1.2, a capsid assembly-defective HBV mutant, in vitro.

We have previously demonstrated that HBcAg dominantly controls the HBV clearance in mice (18). The intrahepatic expression of HBcAg facilitates the development of HBV surface antigen (HBsAg) seroclearance and production of the neutralizing antibody, anti-HBs, in mice. HBcAg serves as a major structural protein of HBV and constitutes the icosahedral capsids. To investigate whether the capsid structure of HBcAg is required for its contributory role in HBV clearance, we took advantage of an HBcAg mutant, HBcY132A, with a tyrosine-to alanine substitution at the 132nd amino acid of HBcAg. This mutation has been reported to prevent HBcAg from self-assembly into the capsid particle while it retains its homodimer formation (5). We used a replication-competent HBV plasmid, pAAV/HBV1.2, as a template to generate the assembly-defective mutant HBcY132A pAAV/HBV1.2 by in situ site-directed mutagenesis without altering other viral genes. To validate the phenotypes of this mutant, we first transfected wild-type (WT) or HBcY132A mutant pAAV/HBV1.2 DNA into HuH-7 cells. The viral transcription, translation, and replication were examined by Northern, Western, and Southern blot analysis, respectively. As shown in Fig. 1A, the mutation of HBcAg did not affect the expression of HBV transcripts, including the 3.5-kb pregenomic RNA and 2.4-, 2.1-, and 0.7-kb RNAs. The levels of two HBV secretory proteins, HBsAg (Fig. 1B) and HBV e antigen (HBeAg) (Fig. 1C), were comparable, as determined by the enzyme immunoassay. In addition, the expression of HBcAg was clearly detected by Western blotting in the cells transfected with WT or HBcY132A mutant DNA (Fig. 1A). However, the formation of HBV capsid particles, revealed by native gel electrophoresis and subsequent Western blotting, was absent in the HBcY132A mutant-transfected cells. The results confirmed that this Y-to-A mutant HBcAg fails to assemble into viral nucleocapsids, which is required for the encapsidated pregenomic RNA to be reverse transcribed and to complete its replication (5). Consistently, the HBcY132A mutant could not support viral replication in vitro (Fig. 1A). Our data confirmed that the mutation of tyrosine to alanine at residue 132 did abolish the ability of HBcAg to self-assemble into capsid particles but did not affect the viral transcription and translation.

Fig 1.

The viral replication and expression of capsid-deficient mutant HBcY132A pAAV/HBV1.2 in vitro. (A) HuH-7 cells were transfected with WT or HBcY132A pAAV/HBV1.2 or mock-transfected (NC), and the viral replication intermediates, including the relaxed circular (RC) and single-stranded (SS) HBV DNAs, were detected by Southern blot analysis (SB) 4 days after transfection. At 2 days after transfection, viral transcripts (3.5-kb pregenomic RNA and 2.1-/2.4-kb and 0.7-kb mRNA), intracellular expression of HBcAg, and levels of secretory HBsAg (n = 2) (B) and HBeAg (n = 2) (C) were determined by Northern blotting (NB), Western blotting (WB), and enzyme immunoassay, respectively. To detect the presence of HBV capsid particles, cytoplasmic lysates were separated by native agarose gel followed by immunoblotting using anti-HBc antibodies 4 days after transfection. The encapsidated HBV DNA was visualized by Southern blotting after native gel electrophoresis. The expressions of GAPDH and β-actin were used as a loading control for NB and WB, respectively.

The capsid structure of HBcAg is crucial for the HBV clearance in vivo.

We have proved that the HBcY132A pAAV/HBV1.2 mutant expressed an assembly-deficient form of HBcAg in vitro. Next, we investigated the importance of capsid formation in the HBV clearance in vivo by hydrodynamic injection of WT pAAV/HBV1.2 or HBcY132A mutant DNA into C57BL/6 mice. The serum levels of HBsAg, a well-known serological marker for HBV infection in humans, were then periodically monitored in these mice. Because HBsAg seroconversion in mice is associated with the clearance of HBV DNA within hepatocytes (15), the disappearance of HBsAg in the sera of these HBV DNA-transfected mice was used as a representative indicator for HBV clearance. Soon after hydrodynamic injection, all the mice receiving WT or HBcY132A mutant DNA developed HBs antigenemia (Fig. 2A). Interestingly, all the mice receiving HBcY132A mutant DNA remained serum HBsAg positive whereas about 80% of those receiving WT DNA became HBsAg negative at 8 weeks postinjection (wpi). The inability to form capsid particles resulted in a prominently higher HBsAg persistence rate in mice.

Fig 2.

Prolonged HBsAg persistence in the mice receiving the HBcY132A pAAV/HBV1.2. C57BL/6 (A) and BALB/c (B) mice were hydrodynamically injected with 10 μg of either WT (●) or HBcY132A (○) pAAV/HBV1.2. Blood samples were collected periodically, and serum levels of HBsAg were determined by enzyme immunoassay (left). The corresponding positive rates of serum HBsAg in these mice were shown (right). The HBsAg positivity is determined as an S/N ratio of ≥2 according to the supplier's criteria. HBsAg persistence rates are significantly different between the two groups of mice (Kaplan-Meier analysis, P = 0.0004 in panel A and P = 0.0001 in panel B).

It has been shown elsewhere that HBc129-140 is an I-A^b-restricted peptide (8, 21). The single-amino-acid substitution of the HBcY132A mutant might destroy this epitope. It is possible that the higher HBsAg persistence rate in HBcY132A-mutant transfected C57BL/6 mice (H2^b) was due to the loss of an I-A^b-restricted epitope of HBcAg. To exclude this possibility, we hydrodynamically injected WT or HBcY132A mutant DNA into the BALB/c mice (H2^d) and monitored their serum levels of HBsAg. Similarly to the result in C57BL/6 mice, the HBsAg-positive rate was significantly higher in the mice receiving HBcY132A mutant DNA than in those receiving WT DNA (Fig. 2B). None of the mice receiving WT DNA remained serum HBsAg positive at 4 wpi. In contrast, serum HBsAg was still detectable in 80% of the mice receiving mutant DNA at 10 wpi. The HBcY132A mutant also resulted in prolonged HBs antigenemia in a mouse strain for which HBc129-140 is not an epitope, indicating that the possible epitope loss of the HBcAg mutant was not the major cause of viral persistence in the mice receiving HBcY132A mutant. Rather, these data suggested that the capsid structure of HBcAg is important for the HBV clearance in vivo. We have previously shown that the intrahepatic expression of HBcAg is correlated with the clearance of HBV in mice using HBeAg/core-null pAAV/HBV1.2, an HBV mutant unable to express HBcAg (18). Here, we further demonstrated that not only the primary structure but also the capsid conformation of HBcAg is required for HBV clearance in vivo.

Intrahepatic expression of full-length but conformationally dysfunctional HBcAg did not contribute to HBV clearance.

The HBV expression in the liver of mice transfected with wild-type or mutant HBV DNAs was examined at different time points after hydrodynamic injection. The viral mRNAs transcribed from the injected HBV DNAs were readily detected in the mice receiving WT or HBcY132A mutant DNA at 3 and 38 days postinjection (dpi) (Fig. 3A). Accordingly, intrahepatic expression of HBcAg and HBsAg was also observed by Western blotting and immunohistochemical staining (Fig. 3A and B). To further verify the inability of HBcAgY132A to form capsid particles in vivo, mouse liver lysates were prepared and subjected to capsid formation assay. Although the denatured forms of WT and HBcY132A mutant HBcAg were comparably detected in the liver of mice, only WT HBcAg assembled into capsid particles (Fig. 4). Capsid-associated viral DNAs reverse transcribed from the encapsidated viral pgRNA were also detected by Southern blotting. Also, viral replication intermediates, including single-stranded (SS) and relaxed circular (RC) forms of HBV DNA, were detected only in the mice receiving WT DNA (Fig. 3A). Consistently, circulating HBV virions could not be detected in the mice receiving HBcY132A mutant DNA (data not shown). Taken together, the HBcAgY132A could not assemble into capsid particles and subsequently hampered the viral replication without affecting the expression levels of viral antigens in mice.

Fig 3.

The intrahepatic viral replication, transcription, and protein expression in the mice hydrodynamically injected with WT or HBcY132A pAAV/HBV1.2. Mice receiving WT or HBcY132A pAAV/HBV1.2 were sacrificed at 3 dpi and 38 dpi to collect liver tissues for the examination of viral replication, transcription, and translation. (A) Intrahepatic HBV plasmid DNAs (Input) as well as viral replication intermediates, including the RC and SS forms of HBV DNAs, were detected using Southern blotting. Viral transcripts were detected by Northern blotting, and GAPDH mRNA was shown as a loading control. Expressions of HBcAg and β-actin (equal loading control) were determined by SDS-PAGE followed by Western blotting. (B) Immunohistochemical staining for the expression of HBcAg and HBsAg in the liver of C57BL/6 mice receiving WT or HBcY132A pAAV/HBV1.2 at 3 dpi (magnification, ×200).

Fig 4.

The failure of HBcY132A mutant to produce HBV nucleocapsids in mice. At 3 dpi, C57BL/6 mice receiving PBS (as a negative control [NC]), WT, HBeAg/core-null, HBcY132A, or HBc175 pAAV/HB1.2 were sacrificed and the livers were collected to prepare cytoplasmic lysates. Individual lysates were separated by native agarose gel electrophoresis and SDS-PAGE to examine the formation of HBV capsid particles and intracellular expression of HBcAg, respectively. Encapsidated HBV DNA was visualized by Southern blotting using DIG-labeled HBV probe. A β-actin blot is shown as a loading control.

Complementation of HBcAgY132A in trans failed to promote HBsAg clearance in mice receiving HBeAg/core-null pAAV/HBV1.2.

To further validate the importance of capsid structure in HBV clearance, we performed a trans-complementation test using an HBV mutant, HBeAg/core-null pAAV/HBV1.2. This mutant cannot express HBcAg and inflamed HBeAg, resulting in prolonged HBV persistence in the mice receiving it (18). The HBeAg/core-null pAAV/HBV1.2 was coinjected with a plasmid encoding either HBcAg or HBcAgY132A with an N-terminal Flag tag into C57BL/6 mice. The capsid-forming ability of Flag-tagged HBcAg and HBcAgY132A is the same as that of their untagged counterparts (Fig. 5B). The intrahepatic expression of tagged WT or mutant HBcAg was also validated in mice (Fig. 5B). As expected, the expression of WT HBcAg drastically decreased the viral persistence rate (Fig. 5A). In contrast, mutant HBcAgY132A coinjection, similar to the result of coinjection with the vector control (pFLAG-CMV2) (18), did not change the persistence rate of HBeAg/core-null mutant. Up to 80% of mice receiving the HBcY132A mutant plasmid remained serum HBsAg positive while the HBsAg-positive rate decreased to nearly 10% in the mice receiving WT HBc plasmid at 8 wpi. This result is consistent with that obtained from the mice in which HBcAgY132A was expressed in the context of the full-length HBV genome. These data further supported the idea that the nucleocapsid assembly of HBcAg is vital to the clearance of HBV.

Fig 5.

The intrahepatic expression of HBcAgY132A did not promote HBsAg clearance in the C57BL/6 mice receiving the HBeAg/core-null pAAV/HBV1.2. Mice were hydrodynamically injected with HBeAg/core-null pAAV/HBV1.2 along with an expression plasmid encoding the WT HBcAg (pFLAG-CMV2/HBc) or mutant HBcAgY132A (pFLAG-CMV2/HBcY132A). (A) Blood samples were collected periodically to determine serum levels of HBsAg (left) and corresponding positive rates of HBsAg in mice (right). The difference in HBsAg-positive rates between the two groups of mice was analyzed by Kaplan-Meier analysis and was shown to be significant (P < 0.0001). (B) Expression of WT or HBcY132A mutant HBcAg from the injected plasmids in the liver of these mice was detected by immunoblotting using anti-HBc antibodies at 1 dpi. Expression of β-actin was shown as a loading control. The lysates were also examined for the formation of capsid particles, as described in Fig. 4.

The HBcY132A mutant impeded the development of host humoral and cellular immunity to HBV in mice.

The antiviral immune responses toward HBV in the mice receiving WT or mutant HBV DNAs were examined. We detected the antibodies against HBcAg (anti-HBc) and HBsAg (anti-HBs) in the sera of C57BL/6 and BALB/c mice after hydrodynamic injection. While all C57BL/6 (Fig. 6A, left) and BALB/c (Fig. 6B, left) mice receiving WT pAAV/HBV1.2 DNA presented anti-HBc antibodies, a clinical HBV infection marker, at 1 wpi, none of the mice receiving the HBeAg/core-null or HBcY132A mutant DNA developed anti-HBc antibodies even after 8 weeks after hydrodynamic injection. Although the HBcY132A mutant encodes a full-length HBcAg with a single-amino-acid change, we could not detect the presence of anti-HBc in these mice. Moreover, most of the mice receiving WT DNA presented anti-HBs antibodies 8 weeks after hydrodynamic injection, with a more rapid kinetics in the BALB/c mice than in the C57BL/6 mice (Fig. 6A and 6B, right). In contrast, anti-HBs-positive rates in both strains of mice receiving HBeAg/core-null or HBcY132A mutant DNA were much lower. Of note, the loss of HBs antigenemia in mice was associated with the anti-HBs seroconversion.

Fig 6.

Impaired antiviral humoral immune responses and IFN-γ response stimulated by the HBcY132A mutant. (A and B) C57BL/6 (A) or BALB/c (B) mice were hydrodynamically injected with WT, HBeAg/core-null, or HBcY132A pAAV/HBV1.2, and blood samples were collected at indicated time points (weeks) after injection to determine the positive rates of serum anti-HBc and anti-HBs antibodies in mice. (C) HBcAg-specific IFN-γ responses of splenocytes isolated from the C57BL/6 mice receiving mock vector, WT, HBeAg/core-null, or HBcY132A pAAV/HBV1.2 (n = 3). Each sample was collected at 10 dpi and was subjected to an ELISPOT assay to measure the frequency of HBcAg-specific IFN-γ-secreting cells. Results were expressed as spot-forming cells (SFCs) per million splenocytes. The asterisk symbolizes statistically significant differences between the WT group and each of the other groups (P < 0.001).

To explore the antiviral cellular immunity in mice, we examined the HBcAg-specific IFN-γ secretion response using an enzyme-linked immunospot (ELISPOT) assay. At 10 dpi, splenocytes isolated from the mice receiving the mock vector, WT, HBeAg/core-null, or HBcY132A pAAV/HBV1.2 were stimulated with the recombinant HBcAg. Then, the frequency of IFN-γ-secreting cells was determined. As shown in Fig. 6C, a significantly high frequency of IFN-γ-secreting cells was observed in the mice receiving WT (613 ± 69 cells per million splenocytes) but not in the mice receiving the mock vector, HBeAg/core-null, or HBcY132A mutant DNA (Fig. 6C) (P < 0.001). The expression of a conformationally defective mutant HBcAg in mice markedly impaired the host in triggering IFN-γ secretion in response to HBcAg. Taken together, our results suggested that the capsid structure is required for HBcAg to elicit effective humoral and cellular immune responses toward HBV that eventually lead to virus clearance. Further studies are warranted to dissect the underlying mechanisms.

DISCUSSION

In this study, we explored the structural requirement of HBcAg to promote HBV clearance in mice hydrodynamically transfected with a capsid-deficient HBV mutant DNA, HBcY132A pAAV/HBV1.2. The elevated HBsAg persistence rate in these mice, but not in those receiving WT HBV DNA, clearly reveals the importance of HBV capsid structure in the viral clearance. We further demonstrated that the capsid structure of HBcAg is required to effectively trigger immune responses that subsequently lead to viral clearance.

The dominant role of HBcAg in the determination of HBV clearance has been shown in a hydrodynamics-based mouse model (18). The intrahepatic expression of HBcAg induced a robust IFN-γ response in mice, facilitating the control of viral infection. Here, we further demonstrated that HBcAg needs to function in the capsid form, but not in the monomer or dimer form, to efficiently mount a protective antiviral immune response. Although the expression of HBcAgY1332A was detected in the liver of mice receiving the assembly-deficient HBV mutant, the majority of these mice failed to elicit appropriate HBV-specific immune responses to eliminate HBs antigenemia. This result suggested that nucleocapsid formation is important for triggering a proper antiviral immune response. Notably, the immune impairment is unlikely to result from the loss of an important epitope which was caused by a single-amino-acid substitution in the mutant HBcAgY132A. Although HBc129-140 serves as an I-A^b-restricted peptide (8, 21), significantly higher HBsAg persistence rates were observed in not only the C57BL/6 (H2^b) (Fig. 2A) but also the BALB/c (H2^d) (Fig. 2B) mice, indicating that the change in this epitope is not responsible for the defective immune responses. Moreover, the possibility that HBcAgY132A exhibits a dominant-negative effect on the HBV-specific immune response can also be ruled out, because it did not promote HBV persistence in mice when it was coexpressed with WT HBV DNA (data not shown).

The capsid assembly from HBcAg is an essential step for HBV replication because the viral pgRNA is reverse transcribed into DNA only when it is packaged into the capsid together with the viral polymerase (26). The HBcY132A pAAV/HBV1.2, which encodes an assembly-deficient mutant HBcAg, was accordingly replication deficient and accompanied by the absence of intracellular nucleocapsids in vitro (Fig. 1A) and in vivo (Fig. 4). In contrast, the viral transcription from the transfected DNA and the expression of viral proteins remained unaffected. Viral replication has been suggested to be unrelated to HBV clearance in mice, as the pol-deficient HBV was cleared at the same rate as the wild type (18). Therefore, the prolonged HBs antigenemia in the mice receiving HBcY132A mutant DNA was not contributed by the defective HBV replication.

Because HBcAg possesses a C-terminal nucleic acid-binding domain which binds to RNA during assembly to form a nucleocapsid, it seems to function in the nucleocapsid form to promote viral clearance. The C terminus of HBcAg nonspecifically binds to cellular or viral RNA during capsid assembly (4, 23, 24), but its DNA binding occurs only when the HBV pol is present, allowing the viral pgRNA to be specifically incorporated into the capsid particle and then reverse transcribed into DNA by the copackaged pol (1, 13). The importance of DNA replication forms within the nucleocapsid as well as the specificity of incorporated RNA can be excluded since the lack of viral polymerase did not affect HBsAg clearance in mice (18). As a result, we proposed that the RNA-containing nucleocapsid, either viral or cellular RNA, contributes to the HBV clearance in vivo.

Capsid structure, however, is not sufficient to support the crucial role of HBcAg in HBV clearance. A previously described HBV mutant, HBc175 pAAV/HBV1.2, which expressed a truncated HBcAg without the C-terminal 10 residues was capable of forming HBV nucleocapsid (Fig. 4), but it did not promote HBV clearance in mice (18). Deletion of a mere 10 residues mitigates the host immune response to HBV without affecting viral replication, expression, and virion production. The release of capsid structure from cells alone is not sufficient to prime the immune response to clear the virus. Surprisingly, no important epitope has yet been identified within this region of HBcAg, excluding the possibility that it directly elicits a strong adaptive immune response against HBV. Besides, the HBcAg-specific humoral immune response remains competent in the HBc175 mutant-receiving mice. Therefore, it is most likely that this mutation influences the antiviral function of innate immunity and subsequently compromises the maturation of adaptive immunity, leading to viral persistence. However, we cannot completely rule out the possibility that the reduced T cell response is attributed to the aberrant processing of HBcAgY132A by the proteasome.

The possible involvement of innate immunity and nucleocapsid structure in the clearance of HBV, as demonstrated by HBc175 pAAV/HBV1.2 and HBcY132A pAAV/HBV1.2 mutants, prompted us to hypothesize that the HBV nucleocapsid can serve as a possible viral PAMP. The polygonal structured capsid is a main feature of most viruses, providing an ideal candidate for a viral PAMP. It is not unprecedented to use nucleocapsid as a PAMP. Viral capsids were recently reported to be a PAMP for adenovirus (6) and retrovirus (20, 22). Although we do not have direct evidence to demonstrate that HBc nucleocapsid can serve as a PAMP to induce innate immunity so far, our studies open a new door to this possibility and warrant further studies.

Whether the innate immunity is activated after HBV infection has aroused a lot of controversy (2). HBV has long been considered a stealth virus since genes involved in the innate response were not induced in HBV-infected chimpanzees (31). A similar conclusion was drawn from human studies (7, 27). However, HBV-induced activation of innate immunity was observed in humans (9, 14), woodchucks (11), and cultured cells (19). Both type I and II IFNs, as well as proinflammatory cytokines such as interleukin-6, are induced in response to HBV infection. Moreover, deficiencies in the innate immunity tampered with the control of HBV infection (10, 12, 37, 38). Using hydrodynamics-based transfection in a series of knockout mice, Yang et al. demonstrated the involvement of innate immunity-related effectors in the HBV clearance, including IFN-α/β receptor 1, IFN-γ, and tumor necrosis factor (TNF) receptor 1 (35). These findings support the idea that innate immunity plays an important role in the control of HBV infection. On the other hand, HBV also evolves to counteract the host innate defense system. The TLR-mediated antiviral activity, for example, was inhibited by HBV (33). Recently, HBV pol and X protein (HBx) were shown to modulate innate immune signaling (16, 29, 30, 36). Thus, a more thorough understanding of the temporal and spatial interaction between HBV and host innate immunity will help us elucidate the discrepancy in current observations. In our recent study, the induction of alpha interferon or beta interferon in the mouse liver after hydrodynamic injection of HBV DNA appeared to be minimal (data not shown), but other innate immune response pathways or the roles of the nonparenchymal cells in the liver will be explored.

How the HBV nucleocapsid interacts with the innate immunity system remains elusive. The crystal structure of the N-terminal assembly domain of HBc has been solved at a fine resolution (34), whereas the flexible C-terminal nucleic acid-binding region of HBc has not been clearly revealed. The fenestrated capsid may allow the C-terminal RNA-binding region of HBcAg to be temporarily exposed outwards for recognition by a cellular PRR, resulting in the subsequent activation of the innate immune response. The C-terminal 10 residues of HBcAg (HBcAg176-185) are likely involved in the engagement of the PRR because a deletion of this region failed to elicit immune responses against HBV. Although the N-terminal assembly domain of HBcAg alone can form a capsid particle resembling that formed from the full-length HBcAg, the stability of the capsid structure is influenced by the C-terminal arginine-rich domain. It has been shown elsewhere that the charge balance between the C-terminal highly basic domain and its associated nucleic acid is important for the capsid stability (17). Moreover, an alteration in the interior C terminus of HBcAg may induce conformational changes of capsid particles (3). These results highlight the involvement of C-terminal HBcAg in the integrity of capsid structure as well as the immunogenicity of the nucleocapsid.

The innate immune system detects viral infection through cellular PRRs. Although the majority of the identified viral PAMPs are nucleic acids, we have shown that without HBcAg, the intracellular viral RNA alone did not result in the viral clearance (18). Therefore, if HBc nucleocapsid can indeed serve as a PAMP, a novel ligand requirement for known PRR or novel PRRs with distinct functions may be involved. In contrast to the nonenveloped adenoviral capsid, which is engaged by its PRR at the cell surface during entry (6), HBV nucleocapsid should be recognized by a yet-unknown intracellular PRR, possibly a cytoplasmic or nuclear PRR. The engagement of nucleocapsid pattern and its corresponding PRR may take place after HBV entry into hepatocytes or internalization of HBV by nonparenchymal liver cells such as Kupffer cells (KCs), natural killer (NK) cells, natural killer T (NKT) cells, and sinusoidal endothelial cells (LSECs). Therefore, the identification of the PRR for HBV nucleocapsid as well as the location of its recognition will facilitate dissecting the molecular mechanism of innate immune activation by HBV and how it affects the maturation of adaptive immunity, constraining viral infection. Additionally, this will help us evaluate the practicability of capsid-disrupting antiviral drugs.

In conclusion, the nucleocapsid structure of HBcAg is required but not solely sufficient for HBV clearance in vivo. The RNA-containing HBV nucleocapsid is crucial to induce an appropriate immune response to control HBV infection. Our findings provide new insights into the role of intact nucleocapsid structure in the innate immunity, which needs further studies to delineate the underlying mechanisms.