Potential conflict of interest: Nothing to report.
Of the 8 genotypes of HBV (genotypes A-H), genotype G is unique in that it has an insertion in the core gene and two stop codons in the precore region preventing the synthesis of hepatitis B e antigen. Most individuals with genotype G are coinfected with other genotypes, typically genotype A. Mice with severe combined immunodeficiency disease carrying human hepatocytes were infected with HBV particles propagated in Huh7 cells in culture. Mice monoinfected with genotype G did not raise detectable HBV DNA in serum, although products of the core gene emerged 4 to 8 weeks after inoculation. When they were superinfected with genotype A at week 10, however, HBV DNA of genotype A developed, which was replaced almost completely by that of genotype G within 10 weeks. Such a rapid takeover was also observed in mice initially infected with genotype A or C and superinfected with genotype G. Similar viral dynamics occurred in mice simultaneously coinfected with genotypes G and A. Takeover was markedly enhanced in mice inoculated with a serum passage containing genotype G with a trace of genotype A. Coinfection of mice with genotypes G and A induced abundant cellular steatosis along with increased fibrosis in the liver, which was not detected in mice monoinfected with genotype A or G. Conclusion: Genotype G can monoinfect chimeric mice at very low levels, and its replication increases maredly when coinfected with other genotypes. Coinfection with genotype G could enhance fibrosis under immunocompromised states. (HEPATOLOGY 2007;45:929–937.)
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
HBV infects an estimated 350 million people worldwide and causes 1 million deaths annually.1 Eight genotypes of HBV have been classified by the sequence divergence in the entire genome exceeding 8% and have been assigned the names A through H in order of discovery.2–6 HBV genotypes have distinct geographic distributions and can influence the severity of liver disease and response to antiviral therapies.7–10 HBV genotypes are further divided into subgenotypes, such as A1/Aa and A2/Ae, B1/Bj and B2/Ba, and C1/Ce and C2/Cs.11–13 These genotypes may influence clinical outcomes of HBV infection.14, 15
HBV genotype G (HBV/G) was first described in 2000 among inhabitants of France and the state of Georgia.5 It has an insertion of 36 base pairs in the core gene and two stop codons in the precore region.5, 16 Despite the inability in encoding hepatitis B e antigen (HBeAg), carriers of HBV/G possess it in serum.5 They are usually coinfected with HBV of other genotypes, most frequently HBV/A, which is responsible for serum HBeAg.17 Coinfection with HBV/C, F, and H has also been reported.18–20 In spite of heavy dependence on other genotypes for replication, HBV/G outgrows them and eventually takes over the great majority of HBV DNA in the circulation.16, 17
Recently, HBV/G DNA in low levels was reported in a German donor of plasmapheresis who had transmitted it to 2 recipients in look-back studies.21 Hence, HBV/G would be able to infect recipients by itself. Furthermore, HBV/G has been detected in 25 of the 104 (24%) French patients coinfected with HIV-1 and HBV and was associated with a high risk of fibrosis at an odds ratio of 12.6.22
Mice with severe combined immunodeficiency disease (SCID) transgenic for the urokinase-type plasminogen activator gene under control of albumin promoter (uPA/SCID mice) have received human hepatocyte transplants.23–25 These mice [hereafter referred to as chimeric (ChiM) mice] have been instrumental in experiments with hepatitis viruses in vivo26, 27 and offer a rare opportunity in portraying the early kinetics of HBV replication,28 without having to resort to the ever-endangered species of chimpanzees. In this study, ChiM mice were monoinfected with HBV/G or coinfected with other genotypes, either simultaneously or in sequence, and followed for circulating HBV/G DNA. It is hoped that the emerging dynamics of HBV DNA will further characterize the dependence of HBV/G on other genotypes and unfold the pathogenicity intrinsic to this parasitic genotype.
ChiM, chimeric mice; HBcrAg, HBV core-related antigen; HBeAg, hepatitis B e antigen; HBsAg, hepatitis B surface antigen; RFLP, restriction fragment length polymorphism; SCID, severe combined immunodeficiency disease.
Patients and Methods
Sera were obtained from 4 patients with chronic hepatitis B. One HBV DNA clone of subgenotype A2 and two of HBV/C2 were recovered from 3 Japanese patients in our recent study.28 Because all HBV DNA clones of HBV/A were classified into subgenotype A2, they will be called HBV/A comprehensively in the present study. The other HBV/A and G clones were obtained from a coinfected Caucasian patient in San Francisco who represented patient 1 in our previous study.17 All the HBV/A or C clones did not have precore or core promoter mutations affecting the expression of HBeAg. The study design conformed to the 1975 Declaration of Helsinki and was approved by the institutional ethics committees. Written informed consent was obtained from each patient.
Plasmid Constructs of HBV DNA and Sequencing.
HBV DNA was extracted from 100 μL of serum using the QIAamp DNA blood kit (Qiagen, GmbH, Hilden, Germany). Four primer sets were designed for amplification of 2 fragments (A and B) covering the entire HBV/G genome. PCR with nested primers was performed with TaKaRa LA Taq polymerase (Takara Biochemicals, Kyoto, Japan) for 35 cycles (30 s at 95°C; 30 s at 60°C; 2 min at 72°C). Primer pairs and protocols for plasmid construction were described in supporting information. As reported previously,28 these fragments were constructed into the pUC19 vector deprived of promoters (Invitrogen Corp., Carlsbad, CA) by digestion with HindIII and EcoRI, resulting in 1.24-fold the HBV genome—just enough to transcribe oversized pregenome and precore messenger RNA. Cloned HBV DNA sequences were confirmed with Prism BigDye (Applied Biosystems, Foster City, CA) using the ABI 3100 automated sequencer. Additionally, HBV DNA spanning the complete genome were amplified in mouse sera, cloned in the pGEM-T Easy Vector, and then sequenced.
Cell Culture and Transfection.
Huh7 cells were transfected with plasmids equivalent to 5 μg of HBV DNA constructs with use of the Fugene 6 transfection reagent (Roche Diagnostics, Indianapolis, IN), and harvested after 3 days in culture. Transfection efficiency was monitored by 0.5 μg of coinfecting reporter plasmids expressing secreted alkaline phosphatase to estimate the latter's enzymatic activity in the culture supernatant.
Determination of HBV Markers.
Hepatitis B surface antigen (HBsAg) and HBeAg were determined via chemiluminescent enzyme immunoassay using commercial assay kits (Fujirebio Inc., Tokyo, Japan). HBV core-related antigens (HBcrAg) were measured in serum using the chemiluminescent enzyme immunoassay described previously.29, 30
Detection and Quantification of Serum HBV DNA.
HBV DNA sequences spanning the S gene were determined via real-time detection PCR according to the method of Abe et al.31 It had a sensitivity of 100 copies/ml (equivalent to 20 IU/ml) on the assay curve obtained with a calibrated World Health Organization standard serum containing HBV of genotype A (kindly provided by Dr. Hiroshi Yoshizawa of Hiroshima University) when 100 μl of the test sample was used. However, in assays for HBV DNA in mouse sera, in which only 10 μl of sample is used, the sensitivity decreased to 1,000 copies/ml (200 IU/ml). For real-time detection PCR specific for HBV/G, 10 μl of DNA sample was amplified in a 25-μl mixture containing 2 × SYBR Green PCR Master Mix (Applied Biosystems) and 2 primers specific for HBV/G: a forward primer (HBVG1620F: ACG TTA CAT GGA AAC CGC CA) and reverse primer (HBVHKR2: AGC CAA AAA GGC CAT ATG GCA) covering the 36–base pair insertion characteristic of this genotype.5, 16 Amplification and detection were performed in the ABI Prism 7700 Sequence Detection System (Applied Biosystems) with an initial activation of UNG at 50°C for 2 minutes, followed by incubation at 95°C for 10 minutes and subsequently, 40 three-step cycles (30 s at 95°C; 30 s at 60°C; 1 min at 72°C) were performed. The standard was prepared on serial dilutions of a known amount of the cloned HBV plasmid of HBV/G. The specificity of 2 primers (HBVG1620F and HBVHKR2) was confirmed in every PCR run via dissociation curve analysis (ABI Prism 7700 dissociation curve software; Applied Biosystems). The sensitivity of detecting HBV/G via real-time detection PCR was 1000 copies/ml (200 IU/ml).
PCR Restriction Fragment Length Polymorphism for Distinguishing HBV DNA of Genotype G from Others.
A novel method for specific determination of HBV/G DNA in the presence of other genotypes has been developed. It involved single-cycle PCR followed by restriction fragment length polymorphism (RFLP) with an endonuclease having the restriction site specific for HBV/G (Fig. 1). PCR was performed with a forward primer (HBV-S190F: GCT CGT GTT ACA GGC GGG) and reverse primer (HBV-S703R: GAA CCA CTG AAC AAA TGG CAC TAG TA) within the S region. To distinguish HBV/G from other genotypes such as HBV/A and C, a portion (5 μl) of amplification products of 510 base pairs was digested with 5 U BstYI (restriction site: RGATCY) at 60°C for 2 hours. Digests were run on electrophoresis in 3.0% agarose gel, stained with ethidium bromide and examined for their sizes under the ultraviolet light. The results were supported by another method (Supplementary Fig. 1).
Inoculation of Chimeric Mice with the Liver Repopulated for Human Hepatocytes.
SCID mice transgenic for the urokinase-type plasminogen activator gene with the liver repopulated for human hepatocytes (chimeric mice) were purchased from Phoenix Bio Co., Ltd. (Hiroshima, Japan). Human serum albumin was measured via ELISA using commercial assay kits (Eiken Chemical Co. Ltd, Tokyo, Japan). They were inoculated with HBV recovered from culture supernatants of Huh7 cells transfected with plasmids constructed with 1.24-fold the HBV genome of genotype HBV/A, C, or G after the method previously reported.28
Liver tissues were fixed in buffered formalin, embedded in paraffin, and stained with hematoxylin-eosin or Masson's trichrome. The fibrosis stage was evaluated by an expert pathologist (S. T.) who was blinded to the nature of inocula.
ChiM Mice Monoinfected with HBV/G.
Two ChiM mice (ChiM92-3 and ChiM184-4) received an inoculum containing approximately 105 copies of HBV/G (G_US1646 strain) and were followed for 12 and 24 weeks, respectively (Fig. 2A,B). HBV DNA remained in undetectable levels (<103 copies/ml) in them both, but they developed low levels of HBcrAg (1 kU/ml) 4 and 8 weeks after inoculation, respectively. Despite absence of detectable HBV DNA in the circulation, therefore, these mice had contracted infection with HBV/G in very low levels. Intrahepatic cccDNA (covalently closed circular DNA) was detected via PCR specific for it,32 and HBV/G DNA was detected in hepatocytes via PCR with type-specific primers33; they attested to infection with HBV/G in them (data not shown).
Superinfection With HBV/A on Mice Infected with HBV/G.
Two chimeric mice with occult infection with HBV/G received 105 copies of HBV/A2 of different strains (A2_JPN to ChiM93-4 and A2_USA to ChiM172-3) 10 weeks after initial inoculation with HBV/G (Fig. 3A,B). They both developed HBV DNA in serum in titers >106 copies/ml at week 17, 7 weeks after superinfection with HBV/A, accompanied by HBcrAg and HBsAg; HBeAg appeared soon thereafter at week 22. HBV DNA and antigens increased, peaked at week 26, and then decreased in exactly the same patterns. HBV/G DNA, which was determined via PCR with type-specific primers, developed 12 weeks after the inoculation at week 22. It increased rapidly, and after the peak, took the same time course as total HBV DNA in serum—it had replaced HBV/A in the two chimeric mice.
Genotypes of HBV in ChiM93-4 were determined at the appearance of HBV DNA (week 17), at peak (week 26), and at the end of observation (week 38) via PCR-RFLP (Fig. 3A). HBV/A accounted for all HBV DNA at week 17. At weeks 26 and 38, however, the vast majority of HBV DNA were of HBV/G with a trace of HBV/A. Thus, HBV/G needed coinfection with HBV/A for active replication, and took it over very swiftly.
Superinfection with HVB/G on Mice Infected with HBV/A.
The chronological order of superinfection was reversed in ChiM92-9 and ChiM124-11 mice (Fig. 4A,B). The mice received 105 copies of HBV/A strains A2_JPN and A2_USA, respectively, and were superinfected with HBV/G (105 copies of G_US1646 strain) 10 weeks thereafter, when HBV/A DNA was elevated to >5 × 107 copies/ml in both strains. Profiles of HBV DNA and antigens in these mice were quite similar but differed from those with A-on-G superinfection (Fig. 3A,B). HBV/A DNA was detected at week 1 in both groups and increased by approximately 2 logs within the next 3 weeks. HBV/G DNA developed within 3 weeks after superinfection with it, much sooner than the 12 weeks in ChiM mice superinfected with 2 genotypes in the reverse order. Three HBV antigens (HBcrAg, HBsAg, and HBeAg) waxed and waned in profiles similar to that of HBV DNA. HBV DNA levels decreased after they had peaked in ChiM92-9 as in A-on-G mice (Fig. 3A,B) by a margin close to log 2; the decrease was less prominent in ChiM124-11 by merely 1 log.
Composition of different genotypes in serum HBV DNA was followed in ChiM92-9 (Fig. 4A). Rapid replacement of HBV/A with HBV/G was obvious in G-on-A superinfection as in A-on-G superinfection (Fig. 3A). The takeover by HBV/G was not complete as in A-on-G superinfection, and HBV/A remained at very low levels throughout the weeks of observation.
Superinfection with HBV/G on Mice Infected with HBV/C.
Similar superinfection with HBV/G was performed on ChiM mice that had been infected with HBV/C2 (Fig. 5A,B). Thus, C_22 and C_AT strains (105 copies) of HBV/C were injected intravenously into ChiM91-21 and ChiM95-11, respectively. They were superinfected with HBV/G (105 copies of G_US1646 strain) at week 10, when HBV DNA stabilized at approximately 109 copies/ml. HBV/G appeared in serum 3 weeks thereafter, at week 13 in both groups, and increased exponentially until weeks 21-23. The time required for an increase in HBV DNA level by 10-fold (log time) was 3.3 weeks in both groups, which was twice as long as the 1.6 weeks in mice with A-on-G and G-on-A superinfections (Figs. 3, 4). Likewise, the takeover of HBV/C by HBV/G in these mice was not as rapid or extensive as in superinfection with HBV/G on HBV/A (Figs. 3A, 4A, 5A). HBV antigens took time courses similar to that of HBV DNA, and they never waned after they had stabilized; however, mice were followed until 26 and 34 weeks.
Simultaneous Coinfection of Mice with HBV/A and HBV/G.
Two ChiM mice (ChiM93-10 and ChiM93-12) received simultaneous inoculation with 105 copies each of HBV/A (A2_JPN strain) and HBV/G (G_US1646 strain). The ChiM93-10 mouse developed HBV/G DNA 17 weeks after inoculation, 9 weeks since HBV/A DNA had increased to >107 copies/ml (Fig. 6A). HBV/G DNA increased to the level of total HBV DNA at week 21, thereby indicating that by then, HBV/G had taken over HBV/A almost completely.
For the reasons unknown, infection with HBV/G was not established in the ChiM93-12 mouse simultaneously coinfected with HBV/A (Fig. 6B), although it was infected with HBV/A in levels by some 2 logs lower (107 copies/ml) than the ChiM93-10 mouse. Serum levels of human albumin in the ChiM93-12 mouse (mean, 2.1 × 106 ng/ml) were much lower than the other chimeric mice used in this study (mean, 4.7 × 106 ng/ml). Thus, a lower extent of repopulation with human hepatocytes may have prohibited active replication of HBV/A. This would be a prerequisite to infection with HBV/G at high levels.
Coinfection of Mice with HBV/A and HBV/G by Inoculation with a Mouse Passage of G-on-A Superinfection.
Three ChiM mice (ChiM169-8, ChiM133-3, and ChiM133-6) received serum from a ChiM92-9 mouse with G-on-A superinfection taken at week 26, when HBV/G had almost replaced HBV/A (Fig. 3A). Profiles of HBV/A and HBV/G, after inoculation with 105 copies of HBV DNA, were similar among the mice (Fig. 7A–C). HBV/G DNA was detected at week 1 in levels comparable to those of total HBV DNA. Despite receiving the inoculation with a mouse passage containing HBV/G, in copies by 5 logs greater than those of HBV/A, HBV/G DNA decreased thereafter and stayed >1 log lower than total HBV DNA until week 7. Since week 4, HBV/G started to increase and replaced HBV/A almost completely until weeks 10-12, and continued to do so through weeks 19-22 of the observation (Fig. 7A).
Cloning and Sequencing HBV DNA in Chimeric Mice Coinfected with HBV/A and HBV/G.
HBV DNA clones from sera of ChiM92-9 sampled at 26 weeks (Fig. 4A) and ChiM169-8 inoculated with serum passage in it (Fig. 7A) included those of HBV/A and G invariably. They confirmed the results of real-time detection PCR and PCR-RFLP and did not possess any mutations in comparison with the original inoculum of either genotype. No recombinations between HBV/A and G were detected, either. At least 5 clones of each genotype were propagated and sequenced in both sera.
Cotransfection of Huh7 Cells with Plasmids Carrying the Core Gene of Genotype A and the Entire Genome of Genotype G.
Huh7 cells were transfected with 2 plasmids that were pcDNA_core clones that expressed the core protein of genotype A2, under the control of cytomegalovirus promoter, and the pUC19/G clone incorporated with 1.24-fold the genome of genotype G. Transfection only with genotype G induced its replication in a weak level (Fig. 8). When Huh7 cells were cotransfected with the genotype G clone and the genotype A core clone, however, the replication was enhanced in a dose-dependent manner.
Liver Pathology of ChiM Mice Infected with HBV/A and/or HBV/G.
Figure 9 shows the histology of liver in representative ChiM mice either simultaneously coinfected with genotypes A and G (viremia of only genotype A in ChiM93-12) or superinfected with genotypes G-on-A (ChiM92-9) and monoinfected with genotype G (ChiM92-3) during 32-39 weeks. HBV infection was demonstrated by double staining for HBcAg and human albumin (Supplementary Fig. 2). The mouse coinfected with genotypes A and G revealed steatosis of hepatocytes with hematoxylin-eosin stain and fibrosis of stage 2 (F2) with Masson's trichrome stain. In contrast, the mice monoinfected with genotype A (ChiM93-12) or G (ChiM92-3) had neither steatosis nor fibrosis. Table 1 summarizes the liver pathology of all autopsied mice. Steatosis in 30%-80% of repopulated human hepatocytes and stage F1-F2 fibrosis were observed in the majority of mice superinfected or coinfected with genotypes G and A or C.
Table 1. Steatosis and Fibrosis in Human Hepatocytes in the Liver of Chimeric Mice Monoinfected or Coinfected with HBV/G
Percentage of human hepatocytes with steatosis. †Simultaneously inoculated with A plus G but became infected with genotype A only (Fig. 6B).
A plus G
A plus G
Using ChiM mice infected with pedigreed HBV DNA in the standardized copy number, we have determined early viral dynamics of HBV/G in detail. Due to constraints on securing ChiM mice with a satisfactory rate of replacement for human hepatocytes (>60%), only 2 or 3 of them were used for each experiment. Concordance of viral dynamics among them, however, would give credence to the reproducibility of obtained results.
HBV/G infected ChiM mice by itself in corroboration with its monoinfection in human beings.21 The replication was very slow, however, and did not elevate serum HBV DNA to levels detectable by the method used (>103 copies/ml). Coinfection with HBV/A enhanced the replication of HBV/G remarkably. HBV/G replicated vividly when coinfected with HBV/C, as well. However, the time required for a 10-fold increase (log time) is 2-fold longer in mice initially infected with HBV/C versus HBV/A (3.3 versus 1.6 weeks). Combined, these results would indicate that HBV/G can thrive at the expense of other genotypes, and coinfection with HBV/A is much more advantageous for its enhanced replication than the other genotypes, including HBV/C. In support of this view, coinfection with HBV/A is frequent in individuals infected with HBV/G.16, 34 Such a heavy dependence of HBV/G on HBV/A does not require recombination between them, because no recombination events occurred in ChiM mice coinfected with them.
The initial replication of HBV/G was much slower than that of HBV/A, even in simultaneous coinfection. This was typically observed in three ChiM mice inoculated with a mouse passage of G-on-A superinfection containing HBV/G in the concentration a few logs higher than that of HBV/A. Despite such an enormous difference in introduced virions, the replication of HBV/A far exceeded that of HBV/G in the initial several weeks. Thereafter, HBV/G caught up with HBV/A, then took it over almost completely. Such a replacement was observed when HBV/G was superinfected on HBV/A, or vice versa.
The mechanism by which genotype G depends on genotype A for replication was pursued in cotransfection experiments in Huh7 cells. Cotransfection with the pcDNA_core clone carrying the core gene of genotype A2 increased the replication of the pUC19/G clone constructed with 1.24-fold the genome of genotype G in a dose-dependent manner (Fig. 8). Hence, trans-complementation with the core protein of genotype A would be required for genotype G to replicate actively. The possibility remains, however, for other viral elements from coinfecting genotypes to enhance the replication of genotype G. Cotransfection of cultured cells with genotype G and others would help clarify how it depends on coinfecting the other genotypes.35
Coinfection with HBV/G may be associated with pathological manifestations. ChiM mice coinfected with HBV/A and HBV/G developed steatosis and fibrosis in the liver not observed in mice monoinfected with either of these genotypes. Very recently, Lacombe et al.22 reported more severe fibrosis in HIV-positive French patients who were infected with HBV/G than the others; they would most likely have been coinfected with other genotypes in undetectable levels. On the basis of clinical and experimental pieces of evidence, it does seem that HBV/G has a strong disease-inducing capacity, which would be operable only when it is coinfected with other genotypes. High levels of HBcrAg in mice with HBV/G (Figs. 3-6) under immunocompromised states would implicate accumulation of the product of the core gene in the fibrosis of patients coinfected with it and HIV. Patients with HIV are infected with HBV at a frequency of 6%-9%, and liver-related deaths happen more often in coinfected patients.36, 37 Fibrosis proceeds faster in patients coinfected with HIV and HBV, as in those with HCV.38, 39 Therapeutic intervention to prevent fibrosis would be required in patients coinfected with HIV and HBV, particularly in HBV/G patients.
In conclusion, the early viral dynamics of HBV/G have been characterized in ChiM mice monoinfected with HBV/G or coinfected with other genotypes. The replication of HBV/G is very slow and depends heavily on coinfection with other genotypes. HBV/G rapidly takes over the other genotypes, though they are indispensable. Infection with HBV/G may induce steatosis and fibrosis in the liver—but again, only in the case of coinfection with other genotypes. However, it is still unclear whether or not such an increased pathogenicity of HBV/G is expressed exclusively in animals and patients with genetic or acquired immune deficiency.