Potential conflict of interest: Nothing to report.
Coinfection with hepatitis B virus (HBV) and hepatitis C virus (HCV) has been associated with severe liver disease and frequent progression to cirrhosis and hepatocellular carcinoma. Clinical evidence suggests reciprocal replicative suppression of the two viruses, or viral interference. However, interactions between HBV and HCV have been difficult to study due to the lack of appropriate model systems. We have established a novel model system to investigate interactions between HBV and HCV. Stable Huh-7 cell lines inducibly replicating HBV were transfected with selectable HCV replicons or infected with cell culture–derived HCV. In this system, both viruses were found to replicate in the same cell without overt interference. Specific inhibition of one virus did not affect the replication and gene expression of the other. Furthermore, cells harboring replicating HBV could be infected with cell culture–derived HCV, arguing against superinfection exclusion. Finally, cells harboring replicating HBV supported efficient production of infectious HCV. Conclusion: HBV and HCV can replicate in the same cell without evidence for direct interference in vitro. Therefore, the viral interference observed in coinfected patients is probably due to indirect mechanisms mediated by innate and/or adaptive host immune responses. These findings provide new insights into the pathogenesis of HBV–HCV coinfection and may contribute to its clinical management in the future. (HEPATOLOGY 2009.)
Hepatitis B virus (HBV) and hepatitis C virus (HCV) are the leading causes of chronic hepatitis, cirrhosis, and hepatocellular carcinoma worldwide. Coinfection with both viruses is common due to shared modes of transmission (reviewed by Alberti et al.1 and Raimondo and Saitta2). Available evidence indicates that fibrosis progression and the development of cirrhosis are more frequent and accelerated in HBV–HCV coinfection as compared with monoinfection by either virus.1, 2 Moreover, a significantly higher incidence of hepatocellular carcinoma and an excess liver-related mortality were noted in coinfection as compared with HBV or HCV infection alone.1–3
The virological and molecular aspects of HBV–HCV coinfection are poorly understood. Although liver disease activity and progression are generally more severe in the presence of double infection, an inverse relationship in the replicative levels of the two viruses has been noted, suggesting direct or indirect (i.e., mediated through host immune responses) viral interference.1, 2 In addition, suppression of HBV replication was found in patients with chronic hepatitis B when they developed acute hepatitis C.4 Likewise, inhibition of HCV replication has been observed in patients with chronic hepatitis C superinfected with HBV.5 Finally, HBV reactivation was observed in some coinfected patients after successful clearance of HCV with pegylated interferon-α (IFN-α) and ribavirin.6, 7 Therefore, while liver disease appears to be enhanced, the two viruses seem to inhibit each other at the replicative level.
Interactions between HBV and HCV have been difficult to study because of the lack of appropriate model systems. The few studies addressing this issue were based on heterologous overexpression of viral proteins and have yielded conflicting results. For example, some studies demonstrated that HCV core inhibits HBV replication,8 whereas others did not.9 Similarly, HCV NS5A was found to enhance10 or inhibit11 HBV replication. Thus, it remains unclear whether there is a direct interference between HBV and HCV.
The aim of this study was to establish an in vitro model system in which both viruses and their interactions could be studied in a replicating context. A tetracycline-regulated gene expression system was used to generate stable Huh-7 cell lines inducibly replicating HBV. These cell lines and control Huh-7 cell lines inducibly expressing the green fluorescent protein (GFP) were transfected with selectable HCV replicons or infected with cell culture–derived HCV (HCVcc). In this system, HBV and HCV were found to replicate without overt interference, suggesting that the reciprocal replicative suppression observed in coinfected patients is due to indirect mechanisms, likely mediated by innate and/or adaptive host immune responses.
pTRE-HBVT contains a 1.05 × HBV genome under control of the tetracycline-controlled transactivator (tTA)-dependent promoter in pTRE2hyg (Clontech, Palo Alto, CA) as described.12 pTRE-hrGFP contains the humanized renilla GFP gene (Dianxing Sun and M.N., unpublished data). The HCV Con1-derived, blasticidin-selectable subgenomic replicon construct pCon1/SG-Bsd(I) was derived from pCon1/SG-Bsd(I)/FlagI.613 (kindly provided by Charles M. Rice, The Rockefeller University, New York, NY). Plasmid pFK-Jc1, coding for the HCV J6/JFH-1 intragenotypic chimera Jc1, has been described.14
Monoclonal antibodies (mAbs) 312 against HBV core and C7-50 against HCV core have been described.15, 16 MAb 11H against HCV NS5A was a gift of Jan Albert Hellings (bioMérieux, Boxtel, The Netherlands). MAb JL-8 against GFP was from Clontech, mAb AC-15 against β-actin was from Sigma (St. Louis, MO), and polyclonal antibody HBcAg Ab-1 against HBV core was from Labvision (Fremont, CA).
Huh-7–derived cell lines constitutively expressing the tTA were established by cotransfection of Huh-7 human hepatocellular carcinoma cells with pUHD15-117 and pCMVNeoBam,18 followed by selection with 200 μg/mL G418. The tTA-dependent luciferase reporter construct pUHD13-317 was used for screening of tightly regulated founder cell lines. H7TA-46.2 and H7TA-61 cells were transfected with pTRE-HBVT or pTRE-hrGFP, followed by double selection with 500 μg/mL G418 and 150 μg/mL hygromycin.
In Vitro Transcription and RNA Electroporation.
In vitro transcription of replicon constructs and RNA electroporation were performed as described.13 Triple-resistant Huh-7–derived cell lines were selected with 500 μg/mL G418, 150 μg/mL hygromycin, and 3 μg/mL blasticidin. In vitro transcription of full-length Jc1 HCV RNA was performed as described.14
RNA Replication Assays.
For quantitative reverse-transcription polymerase chain reaction (RT-PCR), 0.5 μg total cellular RNA was subjected to complementary DNA synthesis using the ThermoScript RT-PCR system (Invitrogen, Carlsbad, CA). HCV RNA and cellular messenger RNA were reverse-transcribed using primer AS165 (5′ TACTCACCGGTTCCGCAGA 3′) and oligo-dT, respectively. Real-time PCR was performed with iQ SYBR Green Supermix (Bio-Rad, Hercules, CA) using primers S66 (5′ ACGCAGAAAGCGTCTAGCCAT 3′) and AS165 on a MyiQ iCycler (Bio-Rad). Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) as an internal reference was amplified using primers GAPDH-S (5′ GAAGGTGAAGGTCGGAGTC 3′) and GAPDH-AS (5′ GAAGATGGTGATGGGATTTC 3′). HCV RNA was quantified by means of comparison with serially diluted in vitro transcripts and normalized to GAPDH.
HCVcc Preparation and 50% Tissue Culture Infective Dose Determination.
Jc1 virus was produced in Huh-7.5 cells19 (kindly provided by Charles M. Rice) as described.14 The 50% tissue culture infective dose was determined as described.20
Northern Blot Analysis.
Denaturing agarose gel electrophoresis and Northern blot were performed according to standard protocols. 32P-labeled complementary DNA fragments representing the entire HCV nonstructural region and a GAPDH probe as a control were employed simultaneously for hybridization.
Southern Blot Analysis.
Southern blot analysis of HBV DNA was performed as described.12 Blots were quantified using ImageQuant TL software (Amersham Pharmacia, Dübendorf, Switzerland).
Immunoblotting was performed as described.16 Native agarose gel electrophoresis and blotting of HBV capsids were performed as described.12
Hepatitis B Surface Antigen Assay.
Hepatitis B surface antigen (HBsAg) in cell culture supernatants was determined by means of an automated assay from Roche Diagnostics (Rotkreuz, Switzerland). Results were expressed as the signal/cutoff ratio normalized to total cellular protein.
Immunofluorescence and Confocal Laser Scanning Microscopy.
Immunofluorescence staining was performed as described.16 Alexa 488–conjugated goat anti-rabbit (Molecular Probes, Eugene, OR) or Cy3-conjugated goat anti-mouse antibodies (Jackson Laboratories, West Grove, PA) were used as secondary antibodies. Slides were examined with a Leica SP5 AOBS confocal laser scanning microscopy (CLSM) system.
IFN-α2a was from Hoffmann-La Roche (Basel, Switzerland). Adefovir (ADV), lamivudine (LAM), telaprevir, and JT-16 were kindly provided by Johan Neyts (Rega Institute for Medical Research, Leuven, Belgium) through the VIRGIL Network of Excellence on Antiviral Drug Resistance. Debio-025 was kindly provided by Grégoire Vuagniaux (Debiopharm, Lausanne, Switzerland). Stock solutions (10 mM) were prepared in dimethyl sulfoxide (DMSO).
Establishment of Huh-7 Cell Lines Replicating HBV and HCV.
Three successive transfection and selection steps allowed the establishment of Huh-7 cells inducibly replicating HBV and constitutively replicating subgenomic HCV RNA (Fig. 1A). In a first step, Huh-7 founder cells constitutively expressing the tTA were established. Screening of approximately 120 G418-resistant clones resulting from the first transfection step allowed the isolation of two clones, designated as H7TA-46.2 and H7TA-61, with an induction capacity of approximately 100-fold in transient transfection assays using a tTA-dependent luciferase reporter (data not shown). In a second step, H7TA-46.2 and H7TA-61 founder cells were stably transfected with a replication-competent, tetracycline-regulated HBV construct, followed by double selection with G418 and hygromycin. In this construct, the tTA-dependent promotor controls transcription of HBV pregenomic RNA, which encodes for core and the polymerase, while preS1, preS2, and S as well as X are constitutively expressed from endogenous HBV promoters. Screening of approximately 100 clones derived from cell line H7TA-46.2 yielded one tightly regulated cell line designated as H7HBV-28.43. Similarly, a well-regulated cell line inducibly replicating and producing infectious HBV was previously derived from cell line H7TA-61.12 A subclone of this cell line, designated as H7HBV-93.1, was used in this study. In addition, H7TA-46.2 and H7TA-61 cells were stably transfected with a tTA-dependent GFP construct, yielding the control cell lines H7GFP-31 and H7GFP-2, respectively. In a third step, H7HBV-93.1 and H7GFP-2 cells were electroporated with an HCV-Con1–derived, blasticidin-selectable replicon harboring the S2204I adaptive change in NS5A. Triple selection with G418, hygromycin, and blasticidin yielded the cell lines H7HBV-93.1-S2204I and H7GFP-2-S2204I, respectively.
H7HBV-93.1-S2204I cells were examined for HBV and HCV replication. For this purpose, cells were cultured in the presence or absence of tetracycline for 5 days, followed by Southern, northern, and western blot analyses. As shown in Fig. 1B, Southern blot analysis using an HBV-specific probe revealed the expected HBV replicative intermediates only in cells cultured in the absence of tetracycline. Similarly, native agarose gel electrophoresis and western blot analysis revealed HBV core only in cells cultured in the absence of tetracycline. By contrast, constitutive HCV replication and protein synthesis could be demonstrated on northern and western blot analyses, respectively, in both H7HBV-93.1-S2204I and H7GFP-2-S2204I cells (Fig. 1C,D). Importantly, HBV replication or GFP expression did not significantly affect HCV RNA replication and protein expression. Similarly, transient transfection of an HCV replicon harboring a luciferase reporter into the different HBV- and GFP-inducible cell lines cultured in the presence or absence of tetracycline did not reveal any inhibition of HCV RNA replication by HBV (data not shown). Taken together, we established stable Huh-7 cell lines allowing the replication of HBV and HCV. In these cell lines, HBV replication did not affect HCV replication—that is, there was no evidence for viral interference.
Subcellular Localization of HBV and HCV Proteins.
Immunofluorescence analyses were performed to assess whether HBV and HCV can replicate in the same cell and whether HBV replication affects the subcellular localization of HCV proteins. In Huh-7 cells harboring HCV replicons, the viral nonstructural proteins accumulate in cytoplasmic dot-like structures that at the ultrastructural level correspond to membranous webs known as viral replication complexes.21 As shown in Fig. 2, CLSM confirmed that H7HBV-93.1-S2204I and H7GFP-2-S2204I cells expressed HBV core and GFP, respectively, only in the absence of tetracycline. As inherently noted with both the tetracycline-regulated gene expression system as well as the HCV replicon system, there was some degree of heterogeneity of protein expression in a given cell line. However, double-label immunofluorescence microscopy of H7HBV-93.1-S2204I cells cultured in the absence of tetracycline revealed that the majority of cells expressed antigens from both viruses (70.6 ± 4.3% of 200 HBV core-positive cells counted in four randomly selected fields were positive for HCV NS5A). Similarly, the majority of H7GFP-2-S2204I cells cultured in the absence of tetracycline were found to be positive for GFP and HCV NS5A (54.3 ± 11.4% of 200 GFP-positive cells counted in four randomly selected fields were positive for HCV NS5A). No colocalization of HBV core and HCV NS5A proteins was found. HBV core was found mainly in the cytoplasm but also in the nucleus of approximately 10% of the cells. NS5A was found in the typical cytoplasmic dot-like distribution. The NS5A fluorescence pattern was indistinguishable in H7HBV-93.1-S2204I and H7GFP-2-S2204I cells cultured in the presence or absence of tetracycline, indicating that HBV replication does not affect the subcellular localization of HCV proteins and replication complex formation. In conclusion, HBV and HCV can replicate in the same Huh-7 cell, and HBV replication does not affect HCV replication complex formation.
IFN-α Responses in Cells Replicating HBV and HCV.
Previous studies have shown an inhibition of IFN-α signaling or IFN-α effector functions by HBV.22 Therefore, we investigated whether the replication of HBV influences the antiviral effect of IFN-α against HCV in our cell lines. It was reported previously that HBV replication in Huh7.93 cells (the cell line from which subclone H7HBV-93.1 was derived) is not inhibited by IFN-α unless very high concentrations (>500 IU/mL) are used.12 By contrast, HCV RNA replication is highly susceptible to inhibition by IFN-α, with IC50 values of 1-2 IU/mL.23 H7HBV-93.1-S2204I cells cultured in the presence or absence of tetracycline were treated with increasing doses of IFN-α for 48 hours, followed by HCV RNA quantitation using real-time RT-PCR. As expected, IFN-α did not inhibit HBV replication at concentrations between 0 and 100 IU/mL (Fig. 3A). Of note, an approximately 50% increase of HBV replicative intermediates was consistently observed at 1 IU/mL, possibly reflecting the stimulatory effect of IFN-α on the tTA-dependent promoter reported previously.24 More importantly, HCV RNA replication was strongly inhibited by IFN-α, independent from HBV replication (Fig. 3B). Similarly, GFP expression in the control cell line H7GFP-2-S2204I did not significantly interfere with the antiviral effect of IFN-α against HCV (Fig. 3B). The data illustrated in Fig. 3 demonstrate the importance of using a model system that allows to control HBV replication in the same cell because of the inherent differences in the permissiveness of individual Huh-7 clones to HCV infection (see below) and replication.25 Taken together, in our model system HBV replication did not interfere with the antiviral effect of IFN-α against HCV.
Specific Inhibition of One Virus Does Not Influence Replication and Gene Expression of the Other.
To further investigate a potential direct interference between HBV and HCV, we treated cells replicating both viruses with selective inhibitors. As shown in Fig. 4A, treatment of H7HBV-93.1-S2204I cells with ADV or LAM for 48 hours efficiently inhibited HBV replication. By contrast, HCV RNA replication was not affected by ADV and LAM treatment of H7HBV-93.1-S2204I or H7GFP-2-S2204I cells (Fig. 4B). These results demonstrate that HBV replication can be efficiently inhibited in the presence of replicating HCV RNA and, furthermore, that inhibition of HBV replication does not affect HCV RNA replication.
To explore whether HCV RNA replication and protein expression influence HBV replication, H7HBV-93.1-S2204I cells were treated with specific HCV inhibitors. As shown in Fig. 5A, the HCV serine protease inhibitor telaprevir (VX-950), the RNA-dependent RNA polymerase inhibitor JT-16, and the cyclophilin inhibitor Debio-025 efficiently inhibit HCV RNA replication in H7HBV-93.1-S2204I or H7GFP-2-S2204I cells cultured in the presence or absence of tetracycline. By contrast, inhibition of HCV RNA replication did not affect HBV replication (Fig. 5B). Interestingly, treatment of H7HBV-93.1-S2204I cells with HCV inhibitors led to a significant (approximately two-fold) increase of HBsAg secretion of Debio-025–treated cells. This increase in HBsAg secretion was independent of HBV replication, consistent with the constitutive production of HBsAg driven by the endogenous promoters for the HBV surface proteins, and may be explained by effects of cyclophilin inhibition on endoplasmic reticulum stress pathways.26, 27 Taken together, in our model system the replication of HBV and HCV could be modulated independently by specific inhibitors. There is no evidence for direct viral interference, as inhibition of one virus does not affect replication of the other.
Superinfection of HBV-Inducible Cell Lines with HCV.
Because we did not observe any significant influence of HBV replication on HCV replication or vice versa, we explored whether cells harboring replicating HBV can be superinfected by HCV using HCVcc. Preliminary experiments revealed that the founder cell lines H7TA-46.2 and H7TA-61, as well as H7HBV-93.1 and H7GFP-2 cells, expressed the HCV entry pathway components CD81 and scavenger receptor type B class I at levels comparable to Huh-7.5 cells,19 a highly permissive Huh-7–derived cell line commonly used for HCVcc infection. Indeed, H7TA-46.2 and H7TA-61 cells were at least as permissive and produced comparable amounts of HCVcc derived from the JFH-1 isolate as well as the intragenotypic chimera Jc1 (data not shown).
H7HBV-93.1 and H7HBV-28.43 cells, as well as the control cell lines H7GFP-2 and H7GFP-31, were infected with Jc1 virus. Forty-eight hours later, cells were stained for HBV and HCV core and the proportion of cells harboring both viruses was counted in randomly selected microscopy fields. As representatively shown in Fig. 6A, H7HBV-93.1 cells were efficiently infected by HCVcc, irrespective of the presence or absence of replicating HBV. Similarly, the control cell line H7GFP-2 was infected independently of GFP expression (Fig. 6B). Analogous results were reproducibly obtained in experiments, including cell lines H7HBV-28.43 and H7GFP-31, as well as in polyclonal pools of H7TA-46.2 and H7TA-61 cells transfected with the tTA-dependent HBV construct and cultured in the absence of tetracycline. As noted previously, there was some clonal variation in the susceptibility to HCVcc infection, but this was not related to the presence or absence of HBV (Fig. 6C and data not shown). We conclude, therefore, that there is no significant superinfection exclusion of HCV by HBV.
HCV Particle Production by Cells Harboring Replicating HBV.
We then investigated the ability of cells harboring replicating HBV to produce infectious HCV. For this purpose, cell lines H7HBV-93.1 and H7HBV-28.43 cultured in the presence or absence of tetracycline were electroporated with Jc1 RNA. As controls, H7GFP-2 and HBVGFP-31 cells were electroporated under the same experimental conditions. HCVcc titers present in the supernatants at 24, 48, and 72 hours after electroporation were determined as the 50% tissue culture infective dose. As shown in Fig. 7, infectious HCVcc was efficiently released from the different cell lines, irrespective of the presence or absence of replicating HBV or of the expression of GFP. These data indicate that neither HBV replication nor HBV surface protein expression interfere with HCV particle production.
Despite the important clinical challenge posed by HBV–HCV coinfection, virtually nothing is known about molecular interactions between HBV and HCV due to the lack of appropriate model systems. Studies performed to date were based on the heterologous overexpression of viral proteins and yielded conflicting results with respect to the effect of HCV core and NS5A proteins on HBV replication.8–11 Here, we describe a novel in vitro model system that allows the analysis of both viruses in a replicating context. Huh-7 cells were used for this purpose because they support HBV replication and virion formation as well as the entire HCV life cycle, including viral entry, RNA replication, and infectious particle release.
In a first step, we generated Huh-7 founder cell lines allowing tightly controlled tetracycline-regulated gene expression. In a second step, these cell lines were stably transfected with a replication-competent HBV construct. In a third step, these cell lines were stably transfected with a subgenomic HCV replicon. Independent HBV-inducible cell lines as well as control cell lines inducibly expressing the GFP were established because significant clonal variation in HCV RNA replication and susceptibility to HCV infection had been noted in the past.14, 20, 25
The three successive transfection and selection steps allowed us to establish Huh-7 cells inducibly replicating HBV and constitutively replicating subgenomic HCV RNA. In these cell lines, it is possible to control the expression of HBV proteins, HBV replication, and (as shown for parental cell line from which H7HBV-93.1 cells were derived12) infectious viral particle formation by the concentration of tetracycline in the culture medium while HCV proteins are expressed and HCV RNA replicated constitutively. Using these cell lines, we demonstrate that HBV and HCV replication are independent and can occur in the same cell. Induction of HBV replication through tetracycline withdrawal did not alter HCV RNA replication, the subcellular localization of HCV proteins, and the formation of cytoplasmic dot-like structures as correlates of viral replication complexes. In addition, specific inhibition of one virus (with LAM or ADV for HBV or with telaprevir, JT-16, or Debio-025 for HCV) did not affect the replication and gene expression of the other. We conclude, therefore, that HBV and HCV can replicate in Huh-7 cells without direct interference. Thus, the two viruses appear to rely on distinct and noncompeting sets of host factors for their replication. In accordance with our in vitro observations, longitudinal studies revealed that the two viruses may replicate independently from one another in some patients, with fluctuations in the serum level of one virus that appear unrelated to viremia of the other.28
Interestingly, HBV replication did not interfere with the antiviral effect of IFN-α against HCV. This is an important observation, because HBV gene products have previously been reported to interfere with IFN-α signaling and effector functions.22 Consistent with our in vitro data, excellent responses of chronic hepatitis C to pegylated IFN-α and ribavirin have been reported recently in HBV–HCV-coinfected patients.6, 7
Our model system also allowed us to investigate HCV infection of Huh-7 cells inducibly replicating HBV. In the HCVcc system used for this purpose, there was no superinfection exclusion of HCV by HBV. Furthermore, cells harboring replicating HBV produced high levels of infectious HCV particles. These results are in agreement with double-label in situ hybridization studies performed in liver biopsies from HBV–HCV-coinfected patients.29
Superinfection exclusion—that is, the ability of an established virus infection to interfere with a secondary virus infection—has been described for a number of viruses, including retroviruses, alphaviruses, vesicular stomatitis virus, and Borna disease virus.30 It is generally restricted to homologous viruses and can occur at various stages of the viral life cycle, thereby favoring entry of newly produced virions into uninfected cells and protecting a primary infecting virus from a competing virus. Among the Flaviviridae, superinfection exclusion has been observed in the pestivirus bovine viral diarrhea virus.30 Moreover, superinfection exclusion of HCV by HCV but not by the related Dengue virus has been reported recently.31, 32 Among the Hepadnaviridae, specific superinfection exclusion of duck hepatitis B virus in primary duck hepatocytes infected with this virus was found to be mediated by the large surface antigen.33 In all cases, infected cells were protected from superinfection by the corresponding homologous virus, whereas nonrelated viruses were able to replicate normally. Therefore, our results demonstrating a lack of superinfection exclusion of HCV by HBV are in agreement with this general concept.
As a limitation, the reverse experiment (the infection of HCV harboring cells by HBV) is not possible in our model system. Indeed, robust infection of cultured cells by HBV remains a challenge, with only few cell types—including HepaRG, primary human, or tupaia hepatocytes—at disposition.34, 35 In this context, HepaRG cells do not appear to be susceptible to efficient HCVcc infection and replication (data not shown) and primary human as well as tupaia cells support only low level HCV replication.
Our study suggests that the reciprocal replicative suppression between HBV and HCV observed in coinfected patients in vivo is not due to a direct antiviral interference but rather due to indirect effects mediated by innate and/or adaptive immune responses. In this context, both HBV and HCV replication are susceptible to antiviral cytokines, including IFN-α, IFN-γ, and TNF-α, that are produced by the infected cell or infiltrating T cells.36 Importantly, further development of our model system by the constitutive expression of major histocompatibility complex class I molecules should allow future investigators to address whether T cell responses against one virus inhibit replication of the other.36
An alternative explanation for the viral interference observed in vivo may relate to host factors that become limiting in human hepatocytes in the liver but not in Huh-7 cells in vitro. Thus, future efforts will be aimed at transfering the described technologies to cells closer to human hepatocytes.
In conclusion, we report a novel cell culture model system allowing the study of HBV–HCV coinfection under highly reproducible conditions in vitro. Our results demonstrate a remarkable lack of direct interference between these two viruses and suggest that indirect effects explain the viral interference observed in coinfected patients in vivo. In addition, our well-characterized cell lines may be valuable for the development and evaluation of new antiviral compounds as well as for studies involving additional hepatotropic viruses such as hepatitis delta virus.
The authors gratefully acknowledge Anja Wahl and Audrey Kennel for excellent technical assistance; Miguel Munoz and Amalio Telenti for support and access to the P3 facility of the Institute of Microbiology of the University of Lausanne; Josef Köck for assistance with HBV Southern blots; and Jan Albert Hellings, Johan Neyts, Charles M. Rice, and Grégoire Vuagniaux for reagents.