The nucleotide sequences of HBV DNA isolates used in this study have been deposited in the GenBank/DDBJ/EMBL databases under accession numbers AB246335–AB246348.
Potential conflict of interest: Nothing to report.
Various genotypes of the hepatitis B virus (HBV) induce liver disease of distinct severity, but the underlying virological differences are not well defined. Huh7 cells were transfected with plasmids carrying 1.24-fold the HBV genome of different genotypes/subgenotypes (2 strains each for Aa/A1, Ae/A2, Ba/B2 and D; 3 each for Bj/B1 and C). HBV DNA levels in cell lysates, determined by Southern hybridization, were the highest for C followed by Bj/Ba and D/Ae (P < .01), and the lowest for Aa (P < .01), whereas in culture media, they were the highest for Bj, distantly followed by Ba/C/D and further by Ae/Aa (P < .01). The intracellular expression of core protein was more than 3-fold lower for Ae/Aa than the others. Hepatitis B e antigen (HBeAg) was excreted in a trend similar to that of HBV DNA with smaller differences. Secretion of hepatitis B surface antigen (HBsAg) was most abundant for Ae followed by Aa, Ba, Bj/C and remotely by D, which was consistent with mRNA levels. Cellular stress determined by the reporter assay for Grp78 promoter was higher for C and Ba than the other genotypes/subgenotypes (P < .01). Severe combined immunodeficiency mice transgenic for urokinase-type plasminogen activator (uPA/SCID), with the liver replaced for human hepatocytes, were inoculated with virions passed in mouse and recovered from culture supernatants. HBV DNA levels in their sera were higher for C than Ae by 2 logs during 4–7 weeks after inoculation. In conclusion, virological differences among HBV genotypes were demonstrated both in vitro and in vivo. These differences may influence HBV infections with distinct genotypes in clinical and epidemiological settings. (HEPATOLOGY 2006;44:915–924.)
Worldwide there are 350 million people persistently infected with hepatitis B virus (HBV) of various genotypes.1 Genotypes of HBV are defined by a sequence divergence >8% over the entire genome made of approximately 3,200 nucleotides (nt).2 Presently, eight genotypes of HBV have been recognized, and they are designated A to H in the order of discovery.3–5
HBV genotypes have different geographic distributions.6–8 Genotype D is ubiquitous and scattered worldwide, while genotype A is prevalent in sub-Saharan Africa, northern America as well as Europe, and genotypes B and C are common in Asia. Genotype E is restricted to Africa, and both genotypes F and H are localized to Central and South America. Because genotype G is infrequent, its epidemiology has not been determined.
Genotypes are further subdivided into subgenotypes. Subgenotype Aa/A1 was originally identified in HBV isolates from South Africa by phylogenetic analysis of preS2/S sequences,9 and later confirmed by the analysis of complete genomes.10, 11 Sugauchi and colleagues12 identified 2 subgenotypes of genotype B. One subgenotype (Bj/B1) is the authentic genotype B indigenous to Japan, whereas the other (Ba/B2) predominates in Asia and has a recombination with genotype C over the precore region and core gene.12, 13 Subgenotypes have also been recognized in genotypes C and D.14, 15
Evidence for the influence of HBV genotypes/subgenotypes on liver disease in acute and chronic infections is increasing.8, 16–18 Because at most two genotypes prevail in a given country, comparison has been restricted between genotypes B and C in Asia as well as A and D in Europe and India,19–21 except in multinational studies where more than two genotypes have been compared.22, 23 Host and environmental differences can confound the differences between genotypes making comparisons very difficult. Therefore a system that eliminates these factors and allows a direct comparison of the influence of HBV genotypes/subgenotypes on viral replication and expression would be ideal.
Since 1986, viral particles have been produced in vitro by transfection of cultured cells with a linear tandem dimer of HBV.24, 25 Virions thus produced are morphologically and virologically indistinguishable from the authentic virion,26 and can infect chimpanzees.27 The minimal length of replication-competent HBV DNA is 1.24-fold genome, which can be transcribed into overlength pregenomic and precore mRNAs.28 Mice with severe combined immunodeficiency, carrying urokinase-type plasminogen activator transgenes controlled by an albumin promoter (uPA/SCID mice),29 have been transplanted with human hepatocytes.30 The graft hepatocytes have morphological and biochemical characteristics identical to human liver31 and can be infected with HBV31–34 and HCV,35 thus providing an ideal small animal model for studying these viruses.
In the present study, by transfection of Huh7 cells and infection of uPA/SCID mice (hereafter referred to as chimeric mice), we evaluated genotype-dependent differences in the intracellular and extracellular expression of HBV DNA and antigens in vitro and in vivo, respectively.
Patients and Methods
Sera were obtained from 14 patients with chronic hepatitis. Demographic and clinical characteristics of the 14 patients and genotypes/subgenotypes of the HBV isolated from them are shown in Table 1. The sera had high HBV DNA levels and contained isolates without the precore G1896A and core promoter A1762T/G1764A mutations. The study design conformed to the 1975 Declaration of Helsinki, and was approved by Ethic Committees of institutions. A written informed consent was obtained from each patient.
Table 1. Demographic, Biochemical, and Virological Characteristics of Patients From Whom HBV Isolates of Distinct Genotypes/Subgenotypes Were Recovered
Country of origin
Abbreviation: LGE, Log genome equivalents.
Plasmid Constructs of HBV DNA and Sequencing.
HBV DNA was extracted from 100 μL of serum using QIAamp DNA blood kits (Qiagen, GmbH, Hilden, Germany). Four primer sets were designed to amplify two fragments (A and B) covering the entire HBV genome. Nested polymerase chain reaction (PCR) was carried out using TaKaRa LA Taq polymerase (Takara Biochemicals, Kyoto, Japan) for 35 cycles (95°C, 30 seconds; 57°C, 30 seconds; 72°C, 2 minutes) (see supplementary information). Amplified fragments were inserted into the pGEM-T Easy Vector (Promega, Madison, WI) and cloned in DH5a competent cells (Toyobo Co. Ltd., Osaka, Japan). At least 5 clones of each fragment were sequenced and the consensus sequence determined. Among the 5 clones, those containing the consensus sequence were identified and used as templates for plasmid construction. To make up the 5′- and 3′-ends of replication-competent construct, fragment C [nt 1413–2815 (nucleotides numbered according to the prototype HBV/C clone with accession no. NC_003977)] was prepared by the fusion PCR technique involving forward primer introduced with an HindIII site and reverse primer with an EcoRI site. Amplified fragment C was cloned in the pGEM-T Easy Vector, and clones having the consensus sequence were selected. Clones for fragment D (nt 2815–1064) and E (nt 1064–2185) bearing the consensus sequence were selected. Fragment C was constructed into the pUC19 vector deprived of promoters (Invitrogen Corp., Carlsbad, CA) by digestion with HindIII and EcoRI. Fragments C, D and E were digested and ligated serially with EcoRI (fragments C and E), EcoO65I (fragments C and D) and EcoT22I (fragments D and E), resulting in 1.24-fold the HBV genome. Cloned HBV DNA sequences were confirmed by using ABI Prism BigDye (Applied Biosystems, Foster City, CA) in the ABI 3100 automated sequencer.
Cell Culture and Transfection.
After 16 hours of culture, Huh7 cells were transfected with 5 μg of DNA construct using the Fugene 6 transfection reagent (Roche Diagnostics, Indianapolis, IN) and harvested 3 days later. Transfection efficiency was measured by cotransfection with 0.5 μg of a reporter plasmid expressing secreted alkaline phosphatase (SEAP) and estimating SEAP enzymatic activity in the culture supernatant. Triple experiments were conducted for each clone.
Determination of HBV Markers.
HBsAg and HBeAg were determined by chemiluminescence with commercial assay kits (Fujirebio Inc., Tokyo, Japan). The product of the HBV core gene (core protein) was determined by enzyme immunoassay using the monoclonal antibody (HB50) that specifically recognizes SPRRR repeats in the arginine-rich domain of core protein.36 The assay is capable of detecting core proteins in viral particles or complexed with antibodies, in addition to free core proteins, because the sample is pretreated to inactivate antibodies and dissociate antigens.
Detection of Extracellular HBV.
The supernatant was collected from each 10-cm dish by centrifugation at 22,000g for 5 min, and a 100-μL portion was adjusted to 6 mM with MgOAc2 and treated with 200 μg/mL DNase I and 100 μg/mL RNase A at 37°C for 3 hours. The reaction was terminated by EDTA at the final concentration of 10 mM, and the mixture was incubated at 65°C for 10 min. HBV DNA was extracted using microspin columns (QIAamp Blood kit, Qiagen K.K, Tokyo, Japan). For real-time detection PCR (RTD-PCR), 10 μL of eluted sample was amplified in a 50-μL mixture containing 2 × TaqMan Universal MasterMix (Applied Biosystems, Foster City, CA), forward primer (HBV-S190F), reverse primer (HBV-S703R) and TaqMan probe (HBSP2G) (details in supplementary information). To avoid the possibility for plasmid DNA remaining in supernatant, Huh7 cells were transfected with plasmid carrying 1.24-fold the HBV DNA with stop-codons in the polymerase gene or plasmid DNA in the absence of transfection reagent. Negative controls processed in parallel never created positive results.
Preparation of RNA.
Transfected cells were lysed by Isogen (Nippon Gene Co. Ltd., Tokyo, Japan). The lysate was supplemented with chloroform, incubated for 15 minutes on ice, and centrifuged at 22,000g for 15 minutes. The aqueous phase was removed and precipitated with isopropanol. RNA was pelleted by centrifugation, washed with ethanol, and dissolved in 50 μL of water.
Southern and Northern Blot Hybridizations.
Southern and Northern blot hybridizations were performed with a full-length probe of each genotype/subgenotype by previous methods.37 No significant differences were observed in the detection between internal control HBV DNA and each probe for all genotypes/subgenotypes.
MTS and Luciferase Assay for Grp78 Promoter.
The pGL3/glucose-regulated protein (Grp78)/−169 reporter plasmid, constructed by subcloning the rat Grp78 promoter subfragment (nt −169 to −29), was generously provided by Amy S. Lee (University of Southern California). Huh7 cells seeded in a 6-well plate at 2 × 105 cells/well were co-transfected with HBV plasmids (0.5 μg), SEAP vector (0.05 μg), and pGL3/Grp78/−169 (0.5 μg), and tested for MTS as well as luciferase 48 hours after transfection. To adjust the number of viable cells, MTS determination was performed by the CellTiter 96 Aqueous One Solution Cell Proliferation Assay (Promega, Madison, WI) in accordance with the manufacturer's instructions. Thereafter in the same well, luciferase activity was determined by LUMAT LB9507 (EG&G Berthold, Bad Wildbad, Germany) and Luciferase Assay System (Promega, Madison, WI) according to the manufacturer's instructions. In addition, the SEAP assay was performed for adjusting the transfection efficiency. Assays were performed at least in triplicate, and the results were expressed as the luciferase activity relative to that of a negative control (mock).
Chimeric Mice With the Liver Repopulated for Human Hepatocytes.
Severe combined immunodeficiency mice transgenic for the urokinase-type plasminogen activator gene (uPA+/+/SCID+/+ mice) with the liver replaced with human hepatocytes38 were purchased from Phoenix Bio Co., Ltd. (Hiroshima, Japan). Human serum albumin was measured by ELISA with commercial assay kits (Bethyl Laboratories Inc., Montgomery, TX). HBV DNA was determined by RTD-PCR as previously reported.39
Freshly prepared liver tissues were snap-frozen in isopentane precooled in liquid nitrogen. Isolated epithelia of bovine cornea were embedded in OCT compound and frozen immediately. Specimens were cut at 5–6 μm by cryostat, mounted on glass slides, air-dried, and fixed in 100% acetone at room temperature for 10 min. Sections were blocked with Antibody Diluent (Dako, Tokyo, Japan), incubated with rabbit anti-HBc (Dako, Tokyo, Japan) at room temperature for 1 hour, washed in phosphate-buffered saline, and then incubated with goat anti-rabbit IgG conjugated with Cy3 (Chemicon International Inc., Temecula, CA) or goat anti-human albumin labeled with FITC (Bethyl Laboratories Inc., Montgomery, TX). Sections were washed with phosphate-buffered saline, and observed in a fluorescent microscope (Eclipse E800M; Nikon, Tokyo, Japan).
Group means were compared by independent Student t test or one-way ANOVA test.
HBV Isolates for Transfection.
HBV isolates of different genotypes/subgenotypes and characteristics of the 14 patients with chronic hepatitis B, from whom these isolates were recovered, are listed in Table 1. At least two isolates for each genotype/subgenotype were tested. None of these isolates possessed mutations for G1896A in the precore region or A1762T/G1764A in the basic core promoter, which may interfere with the expression of HBeAg and efficiency of pregenome encapsidation for replication. The genotypes/subgenotypes of the 14 HBV isolates were determined by the construction of phylogenetic tree, comparing their full genome sequences to 24 sequences from the DNA database, representative of genotypes A to D and subgenotypes of A and B. The 14 HBV isolates cluster with those of corresponding genotypes/subgenotypes (Supplementary Fig. 1).
Intracellular Expression of HBV DNA and Antigens.
Huh7 cells were transfected with a pUC19 vector carrying 1.24-fold the HBV genome. Three days after transfection, they were harvested, lysed with NP-40 and tested for HBV DNA and antigens. The density of single-stranded HBV DNA was compared among different genotypes/subgenotypes by Southern blotting (Fig. 1A). Because the results for the two Ae clones (Ae_US and Ae_JPN) were similar, their mean was set at 1.0, and HBV DNA levels for the other genotypes/subgenotypes were expressed relative to this value. The expression of HBV DNA was the highest for genotype C, followed by two subgenotypes of B and further by A and D (P < .01); it was the lowest for subgenotype Aa.
Figure 1B compares the intracellular expression of core protein in cell lysates among different genotypes/subgenotypes. The expression was the highest for Bj followed by C and Ba. It was relatively lower for D than B and C, still about 3-fold higher for D than Ae and Aa (P < .01).
Extracellular Expression of HBV DNA and Antigens.
After 3 days in culture, supernatants from cells transfected with the 14 HBV strains were compared for HBV DNA, HBeAg and HBsAg (Fig. 2A–C). As in evaluation of intracellular expression, the mean value for Ae was set at 1.0 and used as the reference.
Figure 2A compares HBV DNA levels in culture supernatants. The level of HBV DNA was the highest for Bj, followed by Ba, C and D by a margin of about 3-fold (P < 0.01), and further by Ae and Aa by that of approximately 5-fold (P < .01). These HBV DNA levels were in accord with those able to be immunoprecipitated by anti-HBs and exclusive of naked core particles (supporting Fig. 2). The expression of HBeAg showed a similar trend to HBV DNA with smaller yet significant differences (Fig. 2B).
The expression of HBsAg did not correspond with those of HBV DNA and HBeAg expression (Fig. 2C). HBsAg was expressed in the highest level for Ae, followed by Aa and Ba, and distantly by Bj as well as C (P < .01), and the least for D.
Intracellular Expression of Viral mRNA.
In agreement with the results of HBsAg expression, Northern blot analyses revealed the highest expression levels of preS/S mRNA for Ae similar to those for Aa, followed by Ba, Bj and C, and distantly by D (Fig. 3). In contrast, pregenome/precore mRNA levels were the highest for Bj and C, followed closely by Ba. Genotypes A (Aa/Ae) and D expressed lower pregenome/precore mRNA levels, consistent with low HBV replication for A and D (Figs. 1A, 2A).
To compare the level of cellular stress induced by transfection with distinct HBV genotypes/subgenotypes, a marker for endoplasmic reticulum (ER) stress designated Grp78 was examined; it is one of the best characterized ER chaperon proteins.40, 41 Grp78 promoter activity was highest in cells transfected with HBV/Ba or C, followed by those with D (P < .05), Bj, Ae and Aa (Fig. 4), indicating that HBV/Ba and C could induce higher ER stress in Huh7 cells.
Viral Particles Secreted into Culture Media.
Large Dane particles and small spherical HBsAg particles, precipitated in culture media with anti-HBs, were visualized by immune electron microscopy (details in supporting information). Figure 5 depicts viral particles isolated from the medium of cell cultures transfected with plasmids carrying the Ae genome. Similar electron micrographs were visualized for virions recovered from culture media of cells transfected with the other HBV genotypes/subgenotypes. Using 5-nm colloidal gold as the internal standard, diameters of viral and subviral particles were estimated to be 42 and 22 nm, respectively.
Infection of Chimeric Mice With HBV from Culture Supernatants.
Supernatants of Huh7 cultures harvested 3 days after transfection were condensed in the Amicon concentrator (Millipore SA, Molsheim, France) and injected intravenously into chimeric mice. Sera harvested 3 months after inoculation were diluted to 106 HBV DNA copies/mL. Chimeric mice were inoculated with 100 μL of mouse serum containing 105 HBV DNA copies of subgenotype Ae with the lowest (clone Ae_JPN) or genotype C with the highest (clone C_JPN22) replicative activity (Fig. 1A). Three mice were inoculated with each genotype, and followed weekly for expression of HBV DNA and human albumin in serum (Fig. 6). HBV DNA levels increased 2 weeks after inoculation and continued to rise until 7–8 weeks when they plateaued. HBV DNA titer was higher by 2 logs in the mice inoculated with genotype C than subgenotype Ae. Neither serum levels of human albumin nor the body weight differed between the mice inoculated with Ae and C. Control mice inoculated with plasmids constructed with 1.24-fold the HBV genome of Ae or C did not elicit HBV DNA in serum.
The liver from chimeric mice infected with HBV of subgenotype Ae was examined by immunofluorescent microscopy for hepatitis B core antigen (HBcAg) utilizing anti-HBc labeled with Cy3 (Fig. 7A). The staining for HBcAg was confined to areas where mouse liver had been replaced for human hepatocytes, and the same areas stained for human albumin (Fig. 7B). Co-localization of HBcAg and human hepatocytes was demonstrated by double staining for HBcAg and human albumin (Fig. 7C).
Nucleic acids were extracted from sera of mice infected with subgenotype Ae or genotype C at 12 and 24 weeks after inoculation. The full genome of the HBV isolate was amplified and sequenced. There were no differences in the sequence between isolates obtained from the mouse sera and plasmid constructs used for transfection.
The influence of genotypes/subgenotypes on disease progression and clinical outcome of HBV infection is noted and documented. However, diverse geographical distributions of HBV genotypes/subgenotypes have made it difficult to evaluate these differences, which would be confounded by host and environmental variables. Therefore it would be beneficial to have in vitro and in vivo experimental systems that can control for such variables. In the present study, using in vitro transfection system in Huh7 cells and in vivo uPA/SCID mouse model carrying human hepatocytes, the intracellular and extracellular expression of HBV DNA and antigens was compared among different HBV genotypes/subgenotypes. In order to exclude differences in expression, as a result of variation within strains of the same genotypes/subgenotypes, at least two isolates for each genotype/subgenotype were employed to prepare plasmid constructs with 1.24-fold the HBV genome. HBV isolates containing mutations that affect the expression of HBeAg and viral replication, such as G1896A and A1762T/G1764A, were not used. Both in vitro transfection experiments in Huh7 cells and in vivo infection of uPA/SCID mice with human hepatocytes have demonstrated marked genotype-dependent differences in the expression of HBV DNA and antigens. These differences may contribute to understanding clinical differences among HBV infections with distinct genotypes.
The replication capacity of HBV in transfected Huh7 cells varied among subgenotypes of A (Aa/Ae) and B (Bj/Ba) as well as genotypes C and D (Fig. 1A), with genotype C having the highest replication capacity and subgenotype Aa the lowest. Genotype C is associated with more severe histological liver damage than genotype B.20, 42 It is possible that the intracellular accumulation of HBV DNA and antigens may play a role in inducing liver damage (Table 2). Indeed, the intracellular accumulation of core protein following transfection with genotype C or subgenotype Ba, which is a recombinant of genotype C on genotype B over the precore/core region,12 was higher than those for the other genotypes with the exception of subgenotype Bj (Fig. 1B). Although intracellular HBV DNA levels for Bj were comparable with those for Ba or C, extracellular HBV DNA levels were much higher for subgenotype Bj than Ba or C. The intracellular virion retention was lower for Bj, in reflection of high extracellular HBV DNA expression for this subgenotype. Of note, the Grp78 promoter activity, which is one of the best ER stress markers,40, 41 was the highest for HBV/Ba and C. Increased ER stress (cellular stress) as well as virion retention, observed in Huh7 cells transfected with HBV of either genotype C or subgenotype Ba, could promote inflammation and lead to more severe liver disease than subgenotype Bj.13, 20, 43 On the other hand, a strong tendency of Bj for extracellular virion secretion may endow a high infectious capacity to blood from individuals infected with this subgenotype; it would trigger strong immune responses in hosts. Indeed, such high replication and enhanced secretion of HBV/Bj may be associated with a greater incidence of fulminant hepatitis in individuals infected with subgenotype Bj than genotype C.44 The data on ER stress, however, were obtained by assays of a single reporter gene and without looking into intracellular levels of Grp78 protein. It would be necessary to validate differences in ER stress, among infections with HBV of distinct genotypes, by determining promoter activities of the other genes responsive to ER stress.
Table 2. Summary for Comparison of Intra- and Extracellular Expression of HBV DNA and Antigens, as Well as ER Stress Among Various Genotypes/Subgenotypes
Expression and Stress
Higher than genotypes/subgenotypes without asterisks.
Values relative to negative controls; the other values are relative to that of Ae.
Likewise, intracellular levels of HBV DNA and core protein were higher for genotype D than A, which may increase the activity of liver disease and explain an increased resistance to interferon in patients infected with genotype D.19, 45 Lower replication levels of genotype A than D might be due to imbalance between syntheses of HBsAg and core protein; it would be in favor of HBsAg for genotype A, and core protein for genotype D. The lowest replicative activity of genotype A may explain how HBV/A can evade the immune pressure against it and persist in up to 10% of the adulthood infection.44, 46–48 Higher levels of HBsAg secretion for subgenotypes Ae and Aa than the others may contradict their low replicative activity, but may be an immune escape mechanism. The mechanism needs to be sought for, by which genotype A can direct the synthesis of HBsAg in high levels out of proportion to viral DNA, core protein and HBeAg. High transcription efficiency of preS/S mRNA by subgenotypes Ae and Aa may account for this (Fig. 3). Enhanced hepatocarcinogenic potential of subgenotype Aa in young African adults49 may be related to this high HBsAg expression, too. However, caution must be exercised when extrapolating the results of in vitro experimental models to patients, because the duration of infection or immune responses are not taken into account.
In the present study, chimeric mice were used to compare genotype C and subgenotype Ae. The mice were given an anti-human complement drug to increase the repopulation of mouse liver with human hepatocytes,38 and this was reflected in high levels of human albumin in serum (Fig. 6). Human hepatocytes grafted were successfully infected with viral particles recovered either from culture media of Huh7 cells transfected with plasmid constructs or passages of these particles in mice. Chimeric mice have been infected with HBV recovered from serum31–34 or HepG2 cells transfected with plasmid constructs carrying 1.4-times the HBV genome.34 HBV DNA levels in mouse sera in this study were similar to those reported by Dandri et al.32 and Tsuge et al.,34 but lower than those by Meuleman et al.31, 33 Variables such as zygocities of the uPA gene, as well as sources of hepatocytes and HBV strains for inoculating mice, may make an accurate comparison of reported data difficult. However, by controlling for such variables in the present study, we were able to demonstrate serum levels of HBV DNA by 2 logs higher in mice inoculated with genotype C than subgenotype Ae (Fig. 6). This observation is consistent with the results of transfection studies in Huh7 cells. HBV DNA was expressed in higher levels, either intracellularly (Fig. 1A) or extracellularly (Fig. 2A), by cells transfected with genotype C than subgenotype Ae. Combined, in vitro and in vivo experiments point to a lower replicative activity of subgenotype Ae relative to genotype C.
No mutations were detected in HBV DNA sequences from mice 24 weeks after inoculation in comparison with those of inoculated strains. This is probably ascribable to the lack of immune pressure in severely immunodeficient mice and a low mutation rate of the HBV genome.50 Mutations developing in minor strains may have been missed by the direct sequencing, however.
In conclusion, using both in vitro and in vivo experimental systems, we have been able to demonstrate differences in the expression of HBV markers among distinct genotypes/subgenotypes. These model systems allow comparison among HBV genotypes/subgenotypes free of confounding variables such as host and environmental factors. Furthermore, the lack of immune responses in chimeric mice enables analysis of sheer virological differences. These experimental models may contribute to understanding the influence of genotypes/subgenotypes on clinical outcomes of HBV infection and response to various antiviral treatments.
We thank I. Maruyama and T. Nakamura of PhoenixBio Co. Ltd., Higashi-Hiroshima, Japan for providing chimeric mice with a high replacement for hepatocytes, T. Kimura of Advanced Life Science Institute Inc., Saitama, Japan for examining HBV core protein, Dr. H. Tanaka of Mie University School of Medicine, Mie, Japan for helping to perform IEM, and Drs. C. L. Lai and M. F. Yuen of Queen Mary Hospital, Hong Kong for providing an HBV sample.