Control of hepatitis B virus replication by innate response of HepaRG cells

Authors

  • Julie Lucifora,

    1. INSERM (Institut National de la Santé et de la Recherche Médicale), U871, Lyon, France
    2. Université Lyon I, Lyon, France
    Search for more papers by this author
  • David Durantel,

    1. INSERM (Institut National de la Santé et de la Recherche Médicale), U871, Lyon, France
    2. Université Lyon I, Lyon, France
    3. Hospices Civils de Lyon, Hôtel Dieu Hospital, Lyon, France
    Search for more papers by this author
  • Barbara Testoni,

    1. Department of internal Medecine and Laboratory of Gene Expression, Fondazione A. Cesalpino, University of Rome La Sapienza, Rome, Italy
    2. Laboratoire Associé INSERM, Villejuif, France
    Search for more papers by this author
  • Olivier Hantz,

    1. INSERM (Institut National de la Santé et de la Recherche Médicale), U871, Lyon, France
    2. Université Lyon I, Lyon, France
    Search for more papers by this author
  • Massimo Levrero,

    1. Department of internal Medecine and Laboratory of Gene Expression, Fondazione A. Cesalpino, University of Rome La Sapienza, Rome, Italy
    2. Laboratoire Associé INSERM, Villejuif, France
    Search for more papers by this author
  • Fabien Zoulim

    Corresponding author
    1. INSERM (Institut National de la Santé et de la Recherche Médicale), U871, Lyon, France
    2. Université Lyon I, Lyon, France
    3. Hospices Civils de Lyon, Hôtel Dieu Hospital, Lyon, France
    • INSERM, U871, 151 cours Albert Thomas, 69003 Lyon, France
    Search for more papers by this author
    • fax: (33)-47 2681971.


  • Potential conflict of interest: Nothing to report.

Abstract

Hepatitis B virus (HBV) is currently viewed as a stealth virus that does not elicit innate immunity in vivo. This assumption has not yet been challenged in vitro because of the lack of a relevant cell culture system. The HepaRG cell line, which is physiologically closer to differentiated hepatocytes and permissive to HBV infection, has opened new perspectives in this respect.HBV baculoviruses were used to initiate an HBV replication in both HepG2 and HepaRG cells. To monitor HBV replication, the synthesis of encapsidated DNA, and secretion of hepatitis B surface antigen (HBsAg), was respectively analyzed by southern blot and enzyme-linked immunosorbent assay. The induction of a type I interferon (IFN) response was monitored by targeted quantitative reverse transcription polymerase chain reaction (qRT-PCR), low-density arrays, and functional assays. The invalidation of type I IFN response was obtained by either antibody neutralization or RNA interference. We demonstrate that HBV elicits a strong and specific innate antiviral response that results in a noncytopathic clearance of HBV DNA in HepaRG cells. Challenge experiment showed that transduction with Bac-HBV-WT, but not with control baculoviruses, leads to this antiviral response in HepaRG cells, whereas no antiviral response is observed in HepG2 cells. Cellular gene expression analyses showed that IFN-β and other IFN-stimulated genes were up-regulated in HepG2 and HepaRG cells, but not in cells transduced by control baculoviruses. Interestingly, a rescue of viral replication was observed when IFN-β action was neutralized by antibodies or RNA interference of type I IFN receptor. Conclusion: Our data suggest that a strong HBV replication is able to elicit a type I IFN response in HepaRG-transduced cells. (HEPATOLOGY 2009.)

Hepatitis B virus (HBV) is currently viewed as a noncytopathic virus, and HBV-associated liver damage is thought to be the consequence of a long-lasting cytolytic immune response against infected hepatocytes.1 The outcome of HBV infection, as well as the severity of HBV-induced liver disease, varies widely from one patient to another. In approximately 90% to 95% of adults, exposure to HBV leads to an acute infection, which is rapidly cleared without long-term consequences. The remaining 5% to 10% fail to control viral infection that consequently evolves to chronicity. The latter predisposes patients to severe liver disease, including cirrhosis and hepatocellular carcinoma.2

The intracellular innate antiviral response to infection constitutes an early defense mechanism that aims at stopping replication in already infected cells, as well as limiting the spreading of infection to neighboring cells. This response takes place very early after infection of cells and plays an important stimulatory role for the subsequent development of innate and adaptive immune responses. In many cases, cells mounting this response produce type I interferon (IFN-α, IFN-β) and type III interferons, which are cytokines with pleiotropic functions that mediate both direct and indirect antiviral effects.3–5 The direct antiviral effect of type I IFNs is exerted by a variety of effectors that are expressed by genes whose transcription is directly stimulated by IFNs, in other words, IFN-stimulated genes (ISGs). The production of type I IFNs can be triggered by virus components or replication through cellular sensors that detect the presence of viral RNA, DNA, or proteins.4 The indirect antiviral effect of type I IFNs is attributable to their stimulatory effect on different cells of the innate and adaptive immune response.

Using experimentally infected chimpanzees, microarray analyses suggested that HBV, early in infection, does not modulate host cellular gene transcription and would induce neither innate antiviral response in hepatocytes nor intrahepatic innate immune responses.6 After this study, HBV was qualified as a “stealth virus.”7 However, a study from the same group showed that HBV could be cleared from the livers of infected animals before any detectable adaptive immune response,8 thus suggesting that innate immunity, or antiviral response at the level of infected cells and innate response via specialized cells of immune system (for example, natural killer cells and natural killer T cells), could play an important role. This was eventually demonstrated by a study on two HBV-seronegative blood donors who became HBV DNA positive and were carefully and immediately monitored after the onset of infection. In both patients, natural killer and natural killer T cells were activated before maximal HBV DNA elevation, whereas HBV-specific T cell responses were maximal later when HBV DNA was already declining, thus demonstrating that the innate immune system was able to sense HBV infection.9

The study of antiviral response in HBV-infected cells was hampered by the difficulty of growing the virus in cell culture systems. Only freshly prepared primary human hepatocytes and differentiated HepaRG cells can support a complete HBV life cycle, including early events of infection.10, 11 HepaRG cells are liver bipotent progenitor cells that are able to differentiate into both biliary and hepatocyte-like cells.12 However, the overall replication level in these cells is rather low, with less than 20% of cells infected, which complicates the study of host/pathogen interaction.13 One could hypothesize that the low level of replication might be the consequence of an IFN response. In this case, the virus would be able to trigger a host antiviral response and would be capable of disarming it in only a low percentage of cells. This low percentage of infected cells is an obstacle for studying the potential ability of HBV to elicit an IFN response. Indeed, in other viral models, when a low multiplicity of infection is used, which is likely the case with HBV, it has been documented that an IFN response may occur in only a low percentage (<30%) of infected cells,14 thus complicating its analysis. Moreover, some viruses are particularly efficient at counteracting this IFN response and may therefore render the analysis of IFN response more difficult.15 In this respect, HBV was shown to be very efficient at inhibiting the IFN signaling pathway.16–19 Another technical obstacle for studying the potential ability of HBV to elicit an IFN response is that an inoculation time of 16 hours is required to initiate a strong infection of hepatocytes in vitro.20 This is not compatible with the very early postinfection analysis that would be likely necessary to detect a potentially weak and transient IFN response. To gain data in cell culture system on the potential ability of HBV to elicit and then disarm an IFN response, it appears necessary to initiate a time controlled and high HBV replication in a large number of cells.

One possibility to initiate a time controlled and high intracellular HBV replication in vitro is to use a recombinant baculovirus carrying a full HBV genome to efficiently transduce hepatic cells. This approach led to higher HBV replication levels compared with either transfection of cells or stable cell lines such as HepG2.2.15.21 In a previous study, we have improved this system by generating a new HBV recombinant baculovirus in which the synthesis of pregenomic RNA was driven by a strong mammalian promoter. The initiation of a complete HBV DNA replication cycle, followed by the production of infectious particles, was evidenced in transduced-HepG2 cells.22 However, HepG2 cells are not ideal for studying IFN response, because they are transformed and therefore impaired for many cellular pathways.

The goal of this study was to determine whether HBV could elicit a type I IFN response in the nontransformed/nonneoplastic HepaRG cells that are functional for type I IFN signaling pathway,23 and whether this IFN production could be responsible for the control of HBV replication in this cell line.

Abbreviations

HBsAg, hepatitis B surface antigen; HBV, hepatitis B virus; IFN, interferon; ISG, interferon-stimulated genes; MOI, multiplicity of infection; mRNA, messenger RNA; pfu, plaque-forming units; PT, post-transfection; qRT-PCR, quantitative reverse transcription polymerase chain reaction; shRNA, short hairpin RNA.

Materials and Methods

Cell Culture.

HepG2 (ATCC) cells and HepaRG cells11 were maintained in William's medium (Invitrogen) supplemented with 10% fetal calf serum (Perbio), penicillin/streptomycin 50 U/mL (Invitrogen), glutaMax (Invitrogen) 2mM, insulin bovine 5 μg/mL, and 5 × 10−5 M hydrocortisone hemisuccinate (Roche Diagnostics, Boehringer Mannheim) at 37°C in humidified incubators at 5% CO2. To obtain differentiation of HepaRG, cells were maintained for 2 weeks in standard medium, then for at least 2 more weeks in medium supplemented with 1.8% dimethylsulfoxide (cell culture grade, Sigma).

Baculovirus and Transduction of Mammalian Cells.

Baculoviral constructions (Bac-HBV-WT, Bac-HBV-YPDD, Bac-βGal [Baculovirus containing the β-Galactosidase gene], Bac-GFP [Baculovirus containing the Green Fluorescent Protein]) as well as stock production, titration, and concentration were performed as described previously.22 All the recombinant HBV baculoviruses contain a 1.1× unit-length HBV genome (genotype D, serotype ayw, accession number in GenBank: V01460) that enables the synthesis of pregenomic RNA under the control of chicken beta-actin promoter. Baculoviral transduction of mammalian cells has also been performed as previously described.22

Analysis of Viral DNA.

Purification of HBV DNA from intracellular core particles and analysis/quantification by southern blotting with radioactive probe were performed as previously described.24, 25

Analysis of Gene Expression by Quantitative Reverse Transcription Polymerase Chain Reaction.

Total RNA was extracted from cells with the NucleoSpin RNA II kit according to manufacturer's instructions (Macherey-Nagel). IFN-β gene expression was analyzed by quantitative reverse transcription polymerase chain reaction (qRT-PCR) with the SYBR GreenER Two-Step qRT-PCR Kit for iCycler according to the manufacturer's instructions (Invitrogen). Primers used for the qPCR step are the following : IFN-βFR: 5′GCCGCATTGACCATGTATGAGA3′ and IFN-βRV: 5′GAGATCTTCAGTTTCGGAGGTAAC3 ′.

Analysis of Secreted Type I Interferon.

Three million Huh7.5 cells, which are deficient for retinoic acid inducible gene–induced type I IFN production,26 were transfected with 10 μg pISRE-Luc vector (Stratagene) in a 10-cm diameter dish using Fugene6 transfection reagent according to the manufacturer's instructions (Roche). The pISRE-Luc plasmid (plasmid containing the luciferase gene under the control of the Interferon Stimulated Response Element) expresses luciferase under type I IFN inducible promoter. After 16 hours' incubation with the transfection mixture, cells were trypsinized and reseeded in a 96-well plate at approximately 3 × 104 cells/well in a volume of 100 μL. Six hours later, 100 μL clarified (13,000g, 5 minutes) supernatant, which was collected from transduced cells, was added to the wells. After 24 hours' incubation at 37°C, cells of each well were washed with phosphate-buffered saline and lysed before luciferase activity was monitored using the Renilla Luciferase Assay System (Promega).

Analysis of the IFN Response by Applied Biosystems Low-Density Arrays.

Total RNA was extracted from cells with the NucleoSpin RNA II kit according to the manufacturer's instructions (Macherey-Nagel) and retro-transcribed with the Invitrogen Superscript kit using random examers as template. Two hundred nanograms complementary DNA was loaded in double on the low-density arrays (Applied Bioystems) customized with 95 ISGs probes and 18s probe as a control. The low density arrays (LDAs) were run on a AB 7900HT, and real time PCR data were collected and analyzed through the SDS 2.2 program (Applied Biosystems).

Neutralization of IFN-β.

Neutralization of IFN-β produced in supernatant from cells was performed using a neutralizing anti–IFN-β antibody (ab9662, Abcam) with a final concentration of 0.5 ng/mL.

Construction of HepaRGshINFR Cell Line.

Five different lentiviruses enabling the expression of five different short-hairpin RNAs (shRNA-1: GCCAAGAUUCAGGAAAUUAUU; shRNA-2: CCUUAGUGAUUCAUUCCAUAU; shRNA-3: CGACAUCAUAGAUGACAACUU; shRNA-4: GCUCUCCCGUUUGUCAUUUAU; shRNA-5: GUUGACUCAUUUACACCAUUU) directed against the ifnr1 gene (which codes IFN receptor 1) were obtained as viral stocks from Sigma. HepaRG cells were seeded at 1 × 105 cells in 24-well plates and transduced with the five lentiviruses (multiplicity of infection [MOI] of 5 for each virus). After 16 hours of incubation at 37°C with the transduction mixture, medium containing viruses was replaced by fresh medium containing 20 μg/mL puromycin (InvivoGen) to select for stably transduced cells. Resistant cells were expanded in presence of 40 μg/mL puromycin until the constitution of a frozen stock of cells. The inhibition of the expression of IFNR1 messenger RNA (mRNA) was checked by RT-PCR using the following primers: IFNRI-FR: 5′AGTGTTATGTGGGCTTTGGATGGTTTAAGC3 ′; and IFNRI-RV: 5′TCTGGCTTTCACACAATATACAGTCAGTGG3 ′.

Results

Kinetic of HBV Replication in HepaRG Cells After Transduction with Bac-HBV.

Because the efficiency of transduction was even in both HepG2 and HepaRG cell lines, a direct comparison of the kinetics of HBV replication after transduction with Bac-HBV-WT was possible. Both proliferating and differentiated HepaRG as well as HepG2 cells were transduced at an MOI of 100 plaque-forming units (pfu)/cell with Bac-HBV-WT. Intracellular encapsidated viral DNA was isolated at various times post-transduction (PT) and analyzed by southern blot. HBV DNA was detectable very rapidly after transduction in both HepG2 and HepaRG cells (Fig. 1). Although the amount of HBV intracellular encapsidated DNA was comparable at day 1 PT in both cell lines, kinetics of HBV replication were quite different, because its level decreased slowly in HepG2 cells after 12 days, whereas a rapid and sharp decrease was observed in both proliferating and differentiated HepaRG cells, to become nearly undetectable at day 12 PT. This difference in kinetics could not be attributed to differences in kinetics of clearance of the baculovirus template (data not shown), nor to differential cell death. The fast and noncytolytic elimination of intracellular encapsidated DNA in HepaRG cells suggested that an antiviral response involving specific cellular factors could be involved.

Figure 1.

Kinetic of HBV replication in different cell lines after transduction with an HBV recombinant baculovirus (Bac-HBV-WT). Cells were transduced at an MOI of 100 pfu/cell with Bac-HBV-WT. Encapsidated DNA was extracted at different times PT and analyzed by southern blot. HepaRGp, proliferative HepaRG cells; HepaRGd, differenciated HepaRG cells.

Establishment of an Antiviral State in Bac-HBV-Transduced HepaRG.

To characterize this antiviral state, cells were transduced twice (at day 0 and day 3) with Bac-HBV-WT. As previously shown, the amount of HBV intracellular encapsidated DNA was increased by a second transduction in HepG2 cells,21 whereas it remained unchanged in HepaRG cells (Fig. 2). This result strongly suggests that HepaRG cells were able to mount a sustained antiviral response after the first transduction with Bac-HBV-WT.

Figure 2.

Transduction with HBV recombinant baculoviruses leads to an antiviral state in HepaRG cells but not in HepG2 cells. HepG2 and HepaRG cells were transduced once or twice (days 0 and 3) at an MOI of 100 pfu/cell with Bac-HBV-WT. Encapsidated DNA were extracted at different times PT and analyzed by southern blot. Graph represents quantification of the autoradiogram.

To determine whether this antiviral response in HepaRG cells was attributable to the transgene expression (that is, synthesis of pregenomic RNA, mRNA, and proteins) and replication of HBV genome, or was induced by the baculovirus vector itself, HepG2 or HepaRG cells were first mock-transduced (inoculation with medium only) or transduced with two different control baculoviruses, containing either the β-galactosidase gene (Bac-βgal) or an HBV genome carrying a mutation in the catalytic domain of the polymerase (Bac-HBV-YPDD). The latter baculovirus enables the production of HBV proteins without genome replication.22 Three days after the first transduction, a second transduction was performed with Bac-HBV-WT. In HepG2 cells, HBV intracellular encapsidated DNA was detectable in all experimental conditions (Fig. 3A, upper panel), suggesting that no (or weak) antiviral response was induced in these cells after the first transduction. The apparent differences in signal intensity between mock and Bac-βGal (or Bac-HBV-YPDD) for HepG2 in the shown figure was neither significant nor reproducible. What was highly reproducible was the presence of signal in all three conditions. In contrast, HBV intracellular encapsidated DNA was always detectable in HepaRG cells that were first mock-transduced or transduced with Bac-βGal, but never in those first transduced by Bac-HBV-YPDD (Fig. 3A, bottom panel; see also Fig. 7B). The results reproducibly obtained by southern blot analysis were independently confirmed by the analysis of HBsAg secretion in the same conditions (Fig. 3B). HBsAg was never detected in the supernatant of HepaRG cells that were first transduced with Bac-HBV-YPDD, suggesting a strong IFN-induced degradation of mRNA encoding HBsAg. Altogether those results demonstrate that the antiviral response observed in HepaRG after transduction is mediated by HBV itself and is independent of the baculovirus vector. Moreover, because no reverse transcription step can occur with Bac-HBV-YPDD, this result also indicates that the establishment of an HBV-mediated antiviral state was independent of HBV DNA replication, but rather was attributable to HBV protein or RNA synthesis.

Figure 3.

The antiviral state is attributable to HBV. HepG2 and HepaRG cells were first transduced at an MOI of 100 pfu/cell, either with Bac-HBV-YPDD, Bac-Bgal, or with medium. Three days later, cells were transduced a second time at an MOI of 100 pfu/cell with Bac-HBV-WT. Encapsidated DNA were extracted and analyzed by southern blot (A), and HBsAg was dosed by enzyme-linked immunosorbent assay (B) at different times PT.

Figure 7.

Down-regulation of IFN pathway in HepaRGshIFNR1 cells leads to enhanced HBV replication and prevents the establishment of an HBV-mediated antiviral state. HepaRG and HepaRGshIFNR1 cells were first transduced at an MOI of 100 pfu/cell with either (A) Bac-HBV-WT, (B) Bac-HBV-YPDD, Bac-βgal, or medium. Three days later, cells were transduced or not a second time at an MOI of 100 pfu/cell with Bac-HBV-WT. Encapsidated DNA were extracted at different times PT and analyzed by southern blot.

HBV Elicits IFN-β Production That Subsequently Induces ISGs Expression in Both HepG2 and HepaRG Cells.

Because production of type-I interferons, in particular IFN-β, is one of the first cellular antiviral defenses,4 we first tested whether the ifnβ gene was activated in HepG2 and HepaRG cells (proliferating or differentiated). Cells were transduced with Bac-βGal or Bac-HBV-WT, and qRT-PCR was performed 24 and 48 hours PT. The amount of mRNA encoding IFN-β was strongly increased in cells transduced with Bac-HBV-WT (either HepG2 or HepaRG cells), but not in those transduced with the control baculovirus (Bac-βgal) (Fig. 4A). This indicates that HBV, and not the vector, is responsible for the induction of IFN-β expression in both HepG2 and HepaRG cells. Using a functional assay, we also demonstrated that biologically active type I interferons were produced in the supernatant of Bac-HBV-WT transduced cells (Fig. 4B).

Figure 4.

Transduction with HBV recombinant baculovirus activates the IFN pathway in both HepG2 and HepaRG cells. Cells were transduced at an MOI of 100 pfu/cell with Bac-HBV-WT or with Bac-βGal. Twenty-four and 48 hours after, (A) total RNA was extracted, and IFN-β expression was analyzed by qRT-PCR, and (C) ISG expression was analyzed by qRT-PCR using AB LDA technology customized with 95 ISG and 18s probes. Only selected ISGs are shown (the whole set genes is shown in the Supporting file). (B) Supernatant from transduced cells was also tested for type I IFN activity. All results are expressed in fold change compared with nontransduced cells. HepaRGp, proliferative HepaRG cells; HepaRGd, differenciated HepaRG cells.

Genes activated by type I IFNs are well known, and their expression represents a signature of the IFN pathway.4 To determine whether the pattern of induction after transduction with Bac-HBV-WT was similar to that already described, HepG2 and HepaRG (proliferating or differentiated) cells were transduced with Bac-βGal or Bac-HBV-WT. Expression of 96 ISGs was evaluated by qRT-PCR analyses. Only selected ISGs are shown in Fig. 4C; the results for the whole set of genes are shown in the supplementary file. The control baculovirus did not induce significant changes in the pattern of expression of the genes, thus confirming that the vector is not responsible for the observed IFN response. In contrast, many of the analyzed genes were overexpressed in HepG2, as well as in proliferative and differentiated HepaRG cells 24 and 48 hours after transduction with Bac-HBV-WT.

Many of the overexpressed genes, including viperin, mx1, isg15, oas1, and isg56, have antiviral properties and may be candidate genes responsible for the inhibition of HBV replication in HepaRG. However, these genes were also activated in HepG2 cells in which the replication of HBV was not inhibited as compared with that observed in HepaRG, suggesting that the IFN signaling in these cells do not translate into antiviral effect and confirming the partially nonfunctional IFN pathway in this cell line.27, 28 One gene, CXCL9, was only up-regulated in HepaRG cells, thus representing a signature for this line, whereas three genes represented a signature for HepG2 cells. So far, no correlation between the pattern of gene activation (or repression) and the antiviral response has been evidenced.

Inhibition of Type I IFN Pathway Results in Enhanced HBV Replication in HepaRG Cells.

The next step was to inhibit the IFN pathway in HepaRG cells to determine whether HBV replication could be enhanced. First, proliferating HepaRG cells were transduced with Bac-HBV-WT and immediately treated with a neutralizing anti–IFN-β antibody. Intracellular encapsidated viral DNA was isolated at various times PT and analyzed by southern blot. Results showed that neutralization of IFN-β clearly increased the amount of intracellular encapsidated HBV DNA in proliferating HepaRG (Fig. 5), thus confirming the crucial role of IFN-β in mediating the observed antiviral response in this cell line. It is worth noting that neutralizing anti–IFN-β antibody did not significantly improve HBV replication in HepG2 cells, as shown in Fig, 5B, which presents the fold increase of DNA produced with anti–IFN-β treatment compared with nontreated conditions. This result suggests that the IFN-β produced by HepG2 cells on Bac-HBV transduction, which is functional as shown in Fig, 4B, does not seem to trigger an antiviral response in HepG2.

Figure 5.

Neutralization of IFN-β in supernatant from cells enhances HBV replication. HepaRG, HepG2, or HepaRGshIFNR1 (IFN receptor expression invalidated by RNA interference) cells were transduced at an MOI of 100 pfu/cell with Bac-HBV-WT and immediately treated or not with a neutralizing anti–IFN-β antibody. Treatment was repeated every 3 days when medium was changed. Encapsidated DNA was extracted at different times PT and analyzed by southern blot. (A) Autoradiogram of southern blot obtained with HepaRG cells. (B) Graph showing the ratio between quantity of DNA detected with anti–IFN-β treatment and quantity of DNA detected without treatment.

To further demonstrate the role of type I IFN in inhibiting HBV replication in HepaRG cells, we invalidated the expression of the ifnar1 gene, which codes for subunit 1 of the type I IFN receptor, using the shRNA strategy. The down-regulation of the gene in the resulting cell line, that is, HepaRGshIFNR1, was verified by RT-PCR (Fig. 6A). This down-regulation of the expression of type I IFN receptor is expected to be associated with an inhibition of the amplification of the IFN pathway. To confirm that we significantly invalidated the type I IFN pathway, the expression of ISGs was analyzed by qRT-PCR. We showed that a poly-IC stimulation did not increase the expression of ifn-β, oas1, and isg56 genes (Fig. 6B). Moreover, the levels of mRNA encoding IFN-β, 2',5'-oligoadenylate synthetase (OAs), or Interferon Stimulated Gene 56 (ISG56) were decreased by up to 80% in HepaRGshIFNR1 after transduction with Bac-HBV-WT compared with the parental cell line (Fig. 6C).

Figure 6.

Down-regulation of IFN pathway in HepaRGshIFNR1. HepaRG cells and HepaRGshIFNR1 cells were stimulated either by (A, B) polyIC at 10 μg/mL or (C) by transduction with Bac-HBV-WT at an MOI of 100 pfu/mL. Cells were lysed (A, B) 8 hours after stimulation by polyIC or (C) 24 hours after transduction. Total RNA was extracted and gene expression analyzed by (A) RT-PCR or (B, C) qRT-PCR. Results of qRT-PCR analyses are expressed compared with WT HepaRG cells.

If the type I IFN pathway is crucial for the control of HBV replication, its inhibition should result in increased HBV replication. Both HepaRG and HepaRGshINFR1 were transduced with Bac-HBV-WT and encapsidated HBV DNA extracted at different times PT. The amount of intracellular encapsidated HBV DNA was increased in HepaRGshINFR1 cells, thus demonstrating an inverse correlation between IFN-β and ISGs expression, and HBV replication (Fig. 7A). The inhibition of the antiviral state in HepaRGshINFR1 cells was further evidenced by a double transduction experiment, as performed previously (Fig. 2). Indeed, when two transductions with Bac-HBV-WT were performed at 3-day intervals, this rate of HBV replication was significantly increased in HepaRGshIFNR1 cells as opposed to HepaRG cells (Fig. 7A). It is also worth noting that no additional significant increase in HBV DNA accumulation in HepaRGshIFNR1 cells was observed after IFN-β neutralization (Fig. 5B). This inhibition of the antiviral state in HepaRGshINFR1 was confirmed by successive transductions of cells with different baculoviruses as previously described. Indeed, we showed that, as opposed to wild-type HepaRG cells, HBV intracellular encapsidated DNA could be detected in HepaRGshIFNR1 cells in all conditions and especially when cells were first transduced with Bac-HBV-YPDD (Fig. 7B). Taken together, these results emphasize a clear correlation between the production of IFN-β, inhibition of intracellular HBV accumulation, and installation of an antiviral state in HBV-transduced HepaRG cells.

Discussion

The belief that HBV may be a “stealth virus,” that is, not detected by innate immune defenses, comes from a longitudinal analysis of the activation of cellular genes performed on liver biopsy specimens of three experimentally infected chimpanzees. No data from humans are available to confirm or challenge this result. In chimpanzees, the expression of cellular genes, including those of antiviral cytokines such as IFNα/β, remained apparently unchanged within the liver during the lag phase of infection.6 However, these data should be interpreted with caution, because chimpanzees may not be a completely relevant model, and moreover, one cannot exclude that the inability to detect variation in gene expression may come from the technology used, the stringency of cutoff, as well as the timing of the analysis (in other words, the first gene profiling was performed 2 weeks after the inoculation). The latter point is crucial, because IFN response to viral infection often takes place very rapidly after infection and can be negatively regulated by the virus. Many studies have shown that HBV, when actively replicating, is able to inhibit the IFN signaling pathway.16–19 This capacity to inhibit the IFN signaling pathway, in particular in the context of exogenous IFN exposure, may explain at least partially why chronically infected patients do not respond efficiently to IFN-α treatment.

There are other indirect evidences that innate immunity, that is, antiviral response at the level of infected cells and innate response via specialized cells of the immune system (for example, natural killer, natural killer T cells) may play an important role in controlling HBV infection to some extent. A study showed that HBV clearance could be initiated in the liver of infected chimpanzees before the onset of adaptive immune response.8 In addition, several studies on HBV transgenic mice showed that in animals deficient for type I IFN receptor, protein kinase R, or interferon regulatory factor 1, HBV replicates at higher levels than in control mice.29–31 These observations strongly suggest that type I IFN response contributes to the control of HBV replication in mice. Finally, recent data suggest that the lack of intrahepatic gene induction may be a peculiar feature of the chimpanzee model because an early induction of both innate and adaptive response was observed in two patients with acute HBV infection.9

In vitro data are needed to gain insight in the mechanism by which HBV may induce an innate response, which in turn may control viral replication. Transformed hepatic cell lines (such as HepG2 and Huh7) have limited interest because they do not have fully functional IFN pathways,27, 28 in contrast to primary human hepatocytes and HepaRG that are noncancerous, permissive to some extent to HBV infection,10, 11 and functional for IFN signaling.23 The latter may explain the relatively low replication rate in these cells (< 20% of cells infected).10, 32 This low level of replication, which might result from a cellular antiviral response, complicates the study of host/pathogen interaction.

To gain insight into the potential ability of HBV to elicit then disarm an IFN response, it appeared necessary to initiate a high HBV replication in a large number of cells. In this study, we present evidence that a strong initial HBV infection can induce a type I IFN response, which results in the establishment of an antiviral state that is not overcome by the virus in the nontransformed/nonneoplastic HepaRG cell line. To study this antiviral response, we have used recombinant baculoviruses carrying the HBV genome that are able to transduce a high percentage of cells and trigger a high initial rate of HBV replication. The comparison of the kinetics of HBV replication after cell transduction showed that, in contrast to that observed in HepG2, the presence of intracellular HBV DNA in both proliferating and differentiated HepaRG was transient.

The mechanism of clearance of encapsidated HBV DNA in HepaRG was noncytolytic and correlated with the production of IFN-β and subsequent activation of ISGs. In HepG2 cells, despite the production of IFN-β and activation of ISGs, the amount of intracellular HBV DNA remained stable, thus confirming that type I IFN pathway is not fully functional in this transformed hepatoblastoma-derived cell line, as previously suggested.27, 28, 33 Importantly, we have shown that the establishment of the antiviral state in HepaRG was mainly attributable to the expression of HBV proteins or transcripts, because an HBV mutant with a nonfunctional polymerase was yet able to induce the antiviral state.

The inverse correlation between the activation of type I IFN pathway and inhibition of HBV replication in HepaRG was definitely demonstrated by an experiment in which the action of IFN-β was directly neutralized by antibodies and by experiments with engineered HepaRG in which the type I IFN receptor was down-regulated by RNA interference. In both cases, the rate of HBV replication was greatly increased after transduction. These HepaRGshIFNR1 cells may represent a unique model to study HBV biology and the effect of persistent HBV replication on cell physiology.

Our data demonstrate that, in HepaRG cells, a strong HBV expression and replication induce a potent IFN response that in return restricts infection. It seems that in this model the virus is not able to disarm/counteract the IFN response, despite the demonstrated capacity of some HBV proteins to do so in different experimental conditions.16, 18, 34–36 Further investigations are necessary to determine which viral determinant is responsible for the induction of IFN response in this model and how the virus may overcome it to induce a persistent infection.

Acknowledgements

The authors thank Dr. Isabelle E. Vincent for her advice and critical review of the manuscript.

Ancillary