In vivo proliferation of hepadnavirus-infected hepatocytes induces loss of covalently closed circular DNA in mice

Authors

  • Marc Lutgehetmann,

    1. Department of Medicine, University Medical Center Hamburg-Eppendorf, Hamburg, Germany
    2. Department of Medical Microbiology, Virology and Hygiene, University Medical Center Hamburg-Eppendorf, Hamburg, Germany
    Search for more papers by this author
    • 1

      These authors contributed equally to this work.

  • Tassilo Volz,

    1. Department of Medicine, University Medical Center Hamburg-Eppendorf, Hamburg, Germany
    Search for more papers by this author
    • 1

      These authors contributed equally to this work.

  • Anne Köpke,

    1. Department of Medicine, University Medical Center Hamburg-Eppendorf, Hamburg, Germany
    Search for more papers by this author
  • Tim Broja,

    1. Department of Medicine, University Medical Center Hamburg-Eppendorf, Hamburg, Germany
    Search for more papers by this author
  • Eike Tigges,

    1. Department of Medicine, University Medical Center Hamburg-Eppendorf, Hamburg, Germany
    Search for more papers by this author
  • Ansgar W. Lohse,

    1. Department of Medicine, University Medical Center Hamburg-Eppendorf, Hamburg, Germany
    Search for more papers by this author
  • Eberhard Fuchs,

    1. Clinical Neurobiology Laboratory, German Primate Center, Göttingen, Germany
    Search for more papers by this author
  • John M. Murray,

    1. School of Mathematics and Statistics and Clinical Research, University of New South Wales, Sydney, Australia
    2. National Centre in HIV Epidemiology and Clinical Research, University of New South Wales, Sydney, Australia
    Search for more papers by this author
  • Joerg Petersen,

    1. Department of Medicine, University Medical Center Hamburg-Eppendorf, Hamburg, Germany
    2. Liver Centre Hamburg, Institute for Interdisciplinary Medicine, Asklepios Clinics, St. Georg, Hamburg, Germany
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Maura Dandri

    Corresponding author
    1. Department of Medicine, University Medical Center Hamburg-Eppendorf, Hamburg, Germany
    • Medizinische Klinik und Poliklinik, Zentrum für Innere Medizin, Universitätsklinikum Hamburg-Eppendorf, Martinistr. 52, D-20246 Hamburg, Germany
    Search for more papers by this author
    • These authors contributed equally to this work.

    • fax: (49)-40-7410-57232


  • Potential conflict of interest: Nothing to report.

Abstract

Chronic hepatitis B virus (HBV) infection is maintained by the presence of covalently closed circular DNA (cccDNA), the template of viral transcription and replication. In quiescent hepatocytes, cccDNA is a stable molecule that can persist throughout the hepatocyte lifespan. However, in chronic HBV infection, immunomediated cell injury and compensatory hepatocyte proliferation may favor cccDNA decline and selection of cccDNA-free cells. To investigate the impact of liver regeneration on cccDNA stability and activity in vivo, we used the urokinase-type plasminogen activator (uPA)/severe combined immunodeficiency (SCID) mouse model. Primary tupaia hepatocytes (PTHs) chronically infected with woolly monkey HBV (WM-HBV) were isolated from one highly viremic uPA/SCID chimeric mouse and transplanted into 20 uPA recipients. Expansion of transplanted PTHs and viral load changes were determined by real-time polymerase chain reaction and immunohistochemistry. Transplantation of WM-HBV infected hepatocytes led to an average of 3.8 PTH doublings within 80 days, 75% reduction of virion productivity (relaxed circular DNA/cccDNA), and lower expression levels of pregenomic RNA and hepatitis B core antigen. Remarkably, a median 2-log decline of cccDNA per cell determined during PTH proliferation was due to both dilution of the cccDNA pool among daughter cells and a 0.5-log loss of intrahepatic cccDNA loads (P = 0.02). Intrahepatic viral DNA sequences persisting at the end of the study were mostly present as replicative intermediates and not as integrated virus. Conclusion: Cell division in the setting of liver regeneration and without administration of antiviral drugs induced strong destabilization of the cccDNA reservoir, resulting in cccDNA clearance in the great majority of chronically infected hepatocytes. (Hepatology 2010)

Chronic infection with hepatitis B virus (HBV) affects nearly 400 million people worldwide and is the leading cause of liver cirrhosis and hepatocellular carcinoma. A hallmark of the life cycle of all hepadnaviruses is the formation of stable covalently closed circular DNA (cccDNA) mini-chromosomes associated with histone and nonhistone proteins within the nuclei of infected hepatocytes.1, 2 The presence of cccDNA is responsible for failure of viral clearance and relapse of viral activity after cessation of antiviral therapy with polymerase inhibitors in chronically infected individuals.3

Unlike the provirus DNA of retroviruses, cccDNA does not need to be incorporated into the host genome. It uses the cellular transcriptional machinery to produce four 3′ coterminal viral RNAs necessary for protein production and viral replication, which takes place in the cytoplasm after reverse transcription of the encapsidated pregenomic RNA (3.5 kb). Studies in a duck model indicated that cccDNA molecules can be formed not only from incoming virions infecting the hepatocytes but also from newly synthesized nucleocapsids, which are not enveloped and secreted into the blood but are transported into the nucleus to ensure accumulation and later maintenance of the cccDNA pool.1 Due to this peculiar replication mechanism, cccDNA is not a direct target of current antiviral therapies. Experimental evidence from a woodchuck model and clinical trials shows that in the presence of polymerase inhibitors, the cccDNA pool persisted even when viral production was strongly reduced4, 5 and that long-term antiviral therapy is needed to achieve significant decrease of cccDNA levels.6-8 Considering that cccDNA is very stable in the absence of cell division,5 the development of resistant mutants escaping antiviral therapy may eventually occur. However, it has been proposed that cell division may favor dilution of the cccDNA, so that cccDNA-free cells can be generated while infected cells are forced to divide to compensate for the immunomediated loss of other infected cells.9-11 Notably, studies in the duck model showed that antiviral therapy with polymerase inhibitors induced a greater cccDNA reduction in animals displaying higher hepatocyte proliferation rates.12 A cccDNA decrease was also observed in hepatocytes chronically infected with woodchuck hepatitis virus when cell turnover was induced in vitro by addition of cellular growth factors and viral replication was suppressed by adefovir treatment.4 Furthermore, identification of cccDNA-free woodchuck hepatocytes containing traces of the infection in form of viral integrations indicated that cccDNA clearance may occur without killing the infected cells.9 Due to the narrow host range of HBV and limited availability of infection models, little is known about the impact of liver regeneration on cccDNA stability. Because of a growth stimulus for transplanted hepatocytes that lasts for about 2 to 3 months after cell transplantation, the urokinase-type plasminogen activator (uPA) transgenic uPA mice represent an ideal system to perform studies of cccDNA stability, both during the hepatocyte proliferation phase and during the following quiescent phase, when repopulation of the diseased mouse liver parenchyma is completed.13 Crossing uPA transgenic mice with immunodeficient mice lacking mature T and B lymphocytes permits engraftment and expansion of hepadnavirus-permissive primary hepatocytes.14-17

In this study, we used uPA/severe combined immunodeficiency (SCID) chimeric mice to generate primary tupaia hepatocytes (PTHs) chronically infected with woolly monkey hepatitis B virus (WM-HBV), a hepadnavirus closely related to human HBV but infecting tupaia hepatocytes with higher efficiency. Recipient mice were then repopulated with isolated hepadnavirus-infected hepatocytes to investigate the stability and activity of cccDNA in vivo in the setting of liver regeneration.

Abbreviations:

cccDNA, covalently closed circular DNA; FRET, fluorescence resonance energy transfer; HBcAg, hepatitis B core antigen; HBsAg, hepatitis B surface antigen; HBV, hepatitis B virus; PCNA, proliferating cell nuclear antigen; PCR, polymerase chain reaction; pgRNA, pregenome RNA; preS/S RNA, surface proteins RNA; PTH, primary tupaia hepatocyte; rcDNA, relaxed circular DNA; SCID, severe combined immunodeficiency; uPA, urokinase-type plasminogen activator; WM-HBV, woolly monkey hepatitis B virus.

Materials and Methods

Animals and Mouse Genotyping.

Alb-uPA transgenic mice (Jackson Laboratories, ME) crossed with scid-beige mice (Taconic Farms, Denmark) were housed and maintained under specific pathogen-free conditions in accordance with institutional guidelines under approved protocols. The presence of uPA transgene was determined by way of polymerase chain reaction (PCR)18 and the SCID phenotype was determined by way of immunoglobulin G dot blot. Briefly, mouse serum was spotted onto nitrocellulose membrane (Amersham, CA), incubated with anti-mouse horseradish peroxidase antibodies (1:70,000) (Sigma, Steinheim, Germany) and specific signals were detected by way of chemiluminescence (Pierce, Rockford, IL). Procedures used to establish donor mice harboring PTHs chronically infected with WM-HBV are provided as Supporting Information.

Transplantation Procedures.

One highly viremic mouse chronically infected with WM-HBV was used for liver cell isolation by way of collagenase perfusion. The number of PTHs present in the hepatocyte preparation was determined by real-time PCR using tupaia-specific β-actin primers (forward, AACGAGATGAGATTGGCA; reverse, CAATCCAAGTCCTCGGC; fluorescence resonance energy transfer probes GGTGACAGCAGTCGCAGT-FL and LCRed640-GTTGAAGCGAGCATCCCTAGAGTTCTG-PH) on defined amounts of isolated cells. Serial dilutions of PTH genome equivalents extracted from defined PTH amounts were used as standard. A total of 1 × 106 isolated hepatocytes (both of murine and tupaia origin) were transplanted into 20 uPA+/−/SCID mice by way of intrasplenic injection. Liver specimens removed at the time of sacrifice were snap-frozen in liquid nitrogen for further histological and molecular analyses as described below. All animal experiments were conducted in accordance with the European Communities Council Directive (86/EEC) and were approved by the City of Hamburg and Lower Saxony Federal State Office for Consumer Protection and Food Safety, Germany.

Determination of Liver Repopulation and Viral Loads in Mice.

Viral loads were determined both in serum and livers of infected mice by way of real-time PCR using WM-HBV DNA-specific primers as reported.17 To minimize possible variation in PTH repopulation levels within the liver, three liver samples obtained from distinct regions of each liver were used for nucleic acid isolation. After sample homogenization, DNA and RNA were extracted in parallel using the Master Pure DNA purification kit (Epicentre, Biozym, Germany) and RNeasy RNA purification kit (Qiagen, Hilden, Germany).19 PTH amounts (genome equivalents) and intrahepatic viral loads were determined per nanogram DNA isolated from chimeric livers. After treating 500 ng genomic DNA with 20 U plasmid-safe DNAase I (Epicentre, Biozym), to enrich the cccDNA fraction, intrahepatic cccDNA amounts were determined using WM-HBV cccDNA-specific primers (forward, CTC CCC TCC TGT GCC TTT T; reverse, GCC CAA AAG CCA CCC AAG; FRET probes GGG GTC TCC ATG CAT CTC CAG GTT-FL and LC-Red-AGG TGA AGC GAA GAG CAC ACG GCC-PH). Total WM-HBV DNA and cccDNA copies were normalized for PTH contents as described,16, 17 and cloned WM-HBV DNA was used to establish a standard curve for quantification. Relaxed circular DNA (rcDNA) levels were estimated by subtracting cccDNA amounts from the total WM-HBV DNA. Considering that 1.5 μg DNA was extracted per milligram of liver tissue, PTH genome equivalents and cccDNA loads present per mouse liver were estimated by scaling up values to the total liver weight measured at the time of sacrifice. Viral RNA was reverse-transcribed with Transcriptor (Roche Applied Science, Mannheim, Germany) at 60°C using primers specific for pregenomic RNA (pgRNA), total WM-HBV-RNA, and PTH-specific β-actin sequences.17 Steady-state levels of intrahepatic surface (preS/S) RNAs were determined by subtracting pgRNA amounts from total HBV RNAs (pgRNA + preS/S RNA) estimated in the same liver specimen. pgRNA levels were quantified using WM-HBV pgRNA-specific primers (forward, CACTGTTCACCGCACCATA; reverse, ACCTTCAGTCCCAGTTTG) as well as specific fluorescence resonance energy transfer hybridization probes (TCCTTAGCTTCTTGGGTGGGAAC-FL and LCRed640-ATTTGGAGGATCCTGCTGCTAGAGAATTAGT-PH), as well as the total WM-RNA primers (used for total WM-HBV DNA quantification). WM-hepatitis B surface antigen (HBsAg) was measured by way of enzyme immunoassay (Abbott Axsym HBsAg; Abbott Laboratories, Wiesbaden, Germany).

Separation of rcDNA and cccDNA Molecules From Viral Integrations.

To separate HBV DNA sequences integrated into the host genome from episomal rcDNA and cccDNA molecules, 1 μg DNA was subjected to a 0.8% low melting agarose gel (Plaque GP Agarose, Biozym). DNA ladder (lambda DNA; Fermentas, Hanover, NH) and 100 ng of the same genomic DNA sample were run in parallel and visualized by way of ethidium bromide staining. After removing the high molecular weight DNA band from the gel, the chromosomal DNA was recovered using the GELase Agarose Gel-Digestion Kit (Epicentre, Biozym) according to the manufacturer's recommendations. Purified DNA was then used to determine the presence of integrated viral DNA sequences by way of real-time PCR. Artificial mixtures of genomic DNA and WM-HBV DNA isolated from serum as well as DNA obtained from a human HBV-related hepatocellular carcinoma liver tumor sample and PLC/PRF5 cell line were used as controls.

Immunohistochemistry.

Cryostat sections obtained from distinct regions of transplanted mouse livers were immunostained16, 17 with a human and tupaia-specific cytokeratin 18 monoclonal antibody not cross-reacting with mouse proteins, or with a rabbit hepatitis B core antigen (HBcAg) antiserum (both from DAKO Diagnostika, Hamburg, Germany) or mouse anti-proliferating cell nuclear antigen (PCNA) (Dianova, Hamburg, Germany). Specific signals were then detected using an Envision double-staining kit (Dako, Hamburg, Germany) or Alexa-labeled secondary antibodies (Invitrogen). To determine the fraction of PCNA-positive PTHs in uPA/SCID livers, three slides per animal were analyzed, and 80-200 PTHs were scored per tissue slide, depending on the PTH repopulation levels achieved at different observation times.

Statistics.

The Wilcoxon rank sum test was used for nonparametric pairwise comparisons; P < 0.05 was considered significant. The half-life for cccDNA decay and the doubling time for growth of transplanted PTHs in uPA mouse livers were calculated on linear regression of the log10 of each value determined over time. Significance of correlations was performed using Spearman rank correlation.

Results

Engraftment and Proliferation of Hepadnavirus-Infected Transplanted PTHs.

Real-time PCR analysis revealed that 50% of the hepatocytes isolated from a highly viremic uPA chimeric mouse displaying 4 × 109 WM-HBV DNA copies/mL were of tupaia origin. To estimate repopulation capacities of mouse-derived WM-HBV-infected PTHs, recipient animals were sacrificed at 5, 10, 20, 40, and 80 days after transplantation. Proliferating PTHs were identified in frozen liver sections using tupaia-specific cytokeratin 18 antibodies and PCNA double-staining, while the lack of PTH death was confirmed by performing terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling assays (Supporting Information Fig. 1). As shown in Fig. 1A, transplanted PTHs engrafted mainly as single or cell doublets near the portal vein tracts (day 5). At later time points (Fig. 1B-D), PTH clusters strongly increased in size, demonstrating clonal expansion of the engrafted cells. Although PTH proliferation indexes were highest in the first week after transplantation (median 30% PCNA-positive PTHs [range, 13%-44% (n = 5)] at 5 days), proliferative activity persisted throughout the entire observation time (80 days) displaying median values of 15% (range, 8%-34% [n = 5]) at 10 days, 13% (range, 8%-20% [n = 4]) at 20 days, 8% (range, 4%-12% [n = 3]) at 40 days, and 11% and 16% in the two animals sacrificed at 80 days after cell transplantation, respectively (Fig. 1E).

Figure 1.

Expansion of WM-HBV-infected PTHs in uPA recipients. Mice were sacrificed at 5, 10, 20, 40, and 80 days after transplantation. (A-D) Double immunostaining of frozen liver sections with antibodies specific for tupaia cytokeratin 18 and PCNA. Growth of PTHs (A) 5 days, (B) 20 days, (C) 40 days, and (D) 80 days after transplantation is shown. (E) Median PTH proliferation index determined by counting PCNA-positive PTHs at time of sacrifice. (F) PTH contents per mouse liver were determined by quantifying PTH genome equivalents per nanogram liver DNA and by scaling up values to DNA amounts present in the whole liver mass at time of sacrifice. Arrows indicate PCNA-positive cells.

To estimate the average PTH amounts present in each mouse liver over time, we quantified PTH genome equivalents present per ng liver DNA and scaled up values to the whole liver mass determined for each animal. Liver weight increased 50% after 40 days and 2.2-fold by the end of the study (data not shown), consistent with the increase in mass of these young mice. As shown in Fig. 1F, median 1.9 × 105 PTHs per mouse liver (range, 1-2.6 × 105 [n = 5]) were determined at 5 days, 3.2 × 105 cells (range, 1-4.3 × 105 [n = 5]) at 10 days, 3.7 × 105 PTHs (range, 1.8-4.7 × 105 range [n = 4]) at 20 days, 7 × 105 PTHs (range, 5.7 × 105 to 2.5 × 106 [n = 3]) at 40 days, and 2.8 × 106 PTHs (2.2 and 3.3 × 106 PTHs [n = 2]) at 80 days after transplantation. The median 14-fold increase of PTH number in uPA livers was highly significant (P < 0.0001). Assuming an equal proliferation rate of all engrafted cells, serially transplanted PTHs underwent at least a median of 3.8 cell doublings within 80 days.

Steady-State Levels of Viral Proteins in the Setting of Liver Regeneration.

Double-staining for tupaia cytokeratin 18 and HBcAg showed that all PTHs engrafted in liver sections of mice sacrificed 5 days after transplantation were HBcAg-positive (Fig. 2A,B), and hence chronically infected with WM-HBV. However, HBcAg staining of infected PTHs appeared weaker at later observation times and signals were below detection in PTHs examined 80 days after transplantation, indicating that HBcAg expression levels strongly decreased in regenerating PTHs (Fig. 2C-F). As shown in Table 1, viremia levels were higher in recipient mouse sera at early time points (median 6 and 4 × 105 copies/mL, days 5 and 10, respectively), decreased 1 log thereafter, and were below the lower limit of detection (LLOD) (1 × 104 WM-HBV DNA copies/mL) in one of the two mice analyzed 80 days after transplantation. Serum HBsAg concentrations also declined during expansion of WM-HBV-infected hepatocytes, and they dropped below detection (S/N <2.0 in enzyme-linked immunosorbent assay) in one animal at the end of the study.

Figure 2.

Decrease of HBcAg staining over time in chronically infected PTHs repopulating the uPA mice livers. Serial frozen liver sections of mice sacrificed at (A,B) day 5, (C, D) day 10, and (E, F) day 80 were immunostained with (A,C,E) PTH-specific cytokeratin 18 antibodies or (B,D,F) HBcAg antibodies. Arrows indicate clusters of PTH in mouse livers.

Table 1. Virological Parameters Determined in Serum of Chimeric Mouse Donor, Isolated PTHs, and Recipient Animals
Virological CharacteristicsBaseline*5 days (n = 5)10 days (n = 5)20 days (n = 5)40 days (n = 3)80 days (n = 2)
  • Data are presented as the median (range).

  • Abbreviation: ND, not detected.

  • *

    Values determined either in serum (viremia, HBsAg) of the chimeric mouse donor used for cell isolation or in PTHs isolated from the same animal.

Viremia (WM-HBV DNA/mL serum)2 × 1096 × 105 (5-10 × 105)4 × 105 (2-5 × 105)6 × 104 (4-9 × 104)7 × 104 (6-10 × 104)ND-4 × 104
WM-HBsAg (signal-to-noise; reactive ≥2.0)*2384.3 (4.0-6.5)4.3 (3.8-6.8)2.2 (2.0*-2.5)2.4 (2.2-2.6)2.4 (1.7*-3.1)
rcDNA/PTH1,030185 (14-720)69 (20-128)17 (6-42)2.4 (2-6.7)2 (1.5-2.8)
cccDNA/PTH2.60.6 (0.3-2.3)0.92 (0.2-2.5)0.54 (0.07-1.4)0.05 (0.03-0.06)0.02 (0.02-0.02)
rcDNA/cccDNA41393 (44-325)75 (50-110)52 (12-89)76 (32-131)128 (73-183)
pgRNA/PTH-β-actin (fold expression)11.30.30 (0.04-0.37)0.11 (0.07-0.22)0.04 (0.01-0.06)0.03 (0.003-0.04)0.14 (0.01-0.27)
preS/S RNA/PTH-β-actin (fold expression)12.120.57 (0.43-3.0)1.85 (1.05-3.9)1.68 (0.94-3.16)1.94 (0.35-5.8)0.65 (0.4-0.9)

Changes of Intrahepatic cccDNA Loads After Transplantation of Infected Hepatocytes.

As shown in Table 1, a median of 2.6 copies of cccDNA per PTH was determined within the tupaia-mouse hepatocyte suspension before cell transplantation. Moderate reduction of cccDNA amount per cell was observed within the first 10 days after transplantation (median 0.92 copies cccDNA/PTH [n = 5]). Furthermore, we found that animals harboring higher amounts of cccDNA/PTH at day 10 (2.5 copies) displayed lower PTH/mouse liver (1 × 105), while a greater cccDNA reduction (0.2-0.4 cccDNA/PTH) was achieved in those mice of the same group containing a four-fold higher PTH amounts/liver, suggesting that cccDNA decline correlated with rates of cell expansion. Notably, a median 0.5 and 0.05 cccDNA copies per PTH were measured at day 20 and 40 after transplantation, respectively, and cccDNA loads continued to decline throughout the entire observation time, leading to a 2-log cccDNA reduction per tupaia cell (0.02 cccDNA copies/PTH [P < 0.0001]) in repopulating mouse livers at 80 days after transplantation (Fig. 3A). These results showed that proliferation of WM-HBV infected PTHs (doubling time of 18 days) in uPA mice induced strong reduction of the viral cccDNA template among dividing hepatocytes, leading to an estimation of cccDNA half-life per infected cell of 12 days.

Figure 3.

Hepatocyte proliferation induces significant dilution and loss of intrahepatic cccDNA amounts. (A) log10 reduction of cccDNA copies was estimated per tupaia hepatocyte genome equivalents measured using real-time PCR in the liver of uPA recipients over time (days). (B) log10 decrease of cccDNA contents per mouse liver. Each point represents cccDNA amounts determined in each mouse. Regression lines were used to estimate cccDNA half-life.

To investigate whether cell division also induced significant loss of cccDNA loads in infected livers, cccDNA measurements were scaled for the whole liver content determined in each animal at time of sacrifice. We hypothesized that if cell division occurring in the absence of cell killing (Supporting Information) induced only cccDNA dilution without significant cccDNA loss, then the overall amount of cccDNA present in the entire mouse liver should remain unchanged. However, we determined a 0.5-log reduction of cccDNA loads per mouse liver 80 days after transplantation, compared with the median cccDNA whole liver contents determined at 5 days after transplantation (Fig. 3B), indicating that proliferation of chronically infected PTHs induced not only dilution of the cccDNA among dividing infected cells, but also significant cccDNA loss per mouse liver (P = 0.02), with a cccDNA half-life decay of 34 days. Although we cannot exclude that the original cell suspension harbored some small percentage of cccDNA-negative PTHs, we failed to detect HBcAg-negative PTHs shortly after transplantation (Fig. 2A,B) in all animals analyzed. Therefore, we assumed that the vast majority of PTHs contained at least one cccDNA molecule at engraftment and that the average reduction of 2.6 cccDNA/PTH from donor PTHs to 0.02 cccDNA/PTH determined at day 80 led to cccDNA clearance in a great majority of tupaia hepatocytes.

Reduction of Intrahepatic Virion Productivity.

Quantification of rcDNA amounts revealed that an average of 1,030 rcDNA copies per cell was present in isolated PTHs (Fig. 4A and Table 1). With a median of 185 copies rcDNA/PTH at day 5, 69 at day 10, 17 at day 20, 2.4 at day 40, and 2 at day 80, respectively, we found that expansion of WM-HBV-infected PTHs induced strong reduction of intrahepatic viral loads (Δ3-log rcDNA/PTH [P < 0.0001]). As depicted in Fig. 4B and summarized in Table 1, the intrahepatic viral activity, defined as rcDNA copies determined per cccDNA molecule, dropped from 413 rcDNA/cccDNA estimated before transplantation to 93 copies rcDNA/cccDNA determined at day 5. Median intrahepatic viral activity was 75 rcDNA/cccDNA at day 10, 52 at day 20, and 76 at day 40. The higher viral activity determined in one of the two repopulated animals analyzed at 80 days after transplantation (183 rcDNA/cccDNA) (Table 1) suggests that suppression of cccDNA activity may have relented toward the end of the regeneration process.

Figure 4.

Reduction of intrahepatic viral loads in proliferating PTHs after transplantation (Tx). (A) Changes of intrahepatic rcDNA amounts normalized per PTH were determined using real-time PCR by subtracting cccDNA copies to total HBV DNA loads. Each point represents rcDNA amounts determined in each mouse. (B) Virion productivity estimated as rcDNA copies per cccDNA amounts determined before hepatocyte transplantation (time 0) and at different time points in the liver of recipients animals analyzed. (C) Log median levels of pgRNA and preS/S RNA normalized per PTH-specific β-actin RNA levels determined both before transplantation of isolated PTHs and in each mouse liver at 5, 10, 20, 40, and 80 days after transplantation.

To investigate whether reduction of rcDNA replicative intermediates correlated with a similar reduction of the steady-state levels of pgRNA transcription, pgRNA levels were determined using primers designed to amplify specifically the pgRNA sequences and not the subgenomic viral RNAs. After normalization of pgRNA levels to PTH-specific β-actin expression levels (Fig. 4C), we found that levels of pregenomic viral RNA were significantly lower within the first week after transplantation (1.6-log reduction) and that transcriptional activity continued to decrease (Δ1 log) between day 5 and day 40. The very good correlation (R2 = 0.77) determined between levels of rcDNA and pgRNA in all samples analyzed indicated that suppression of virion productivity was due to lower steady-state levels of pgRNA. In agreement with the rcDNA contents determined at the end of the study, increase of pgRNA was observed in one of the two animals analyzed 80 days after serial transplantation, again suggesting that transcriptional activity, or stability of the transcribed pgRNA, may be restored when hepatocyte expansion relents.

As shown in Fig. 4C, steady-state levels of subgenomic preS/S RNA also decreased during liver regeneration as the average number of cccDNA copies per cell dropped. Nevertheless, the decline of subgenomic viral RNAs was less pronounced (Δ1 log preS/S RNA) compared with the pgRNA reduction determined over time (Δ2.5 log), suggesting that HBsAg production may be less affected during liver regeneration than the replicative pathway.

Persistence of Chronic Infection in PTHs Repopulating uPA Livers Is Not Due to Occurrence of Viral Integrations.

To examine whether WM-HBV DNA sequences measured in recipient uPA livers originated not only from persisting viral replicative intermediates but also from sequences integrated into the host genome, we performed gel electrophoresis to separate the low molecular weight rcDNA and cccDNA forms from the high molecular weight genomic DNA, possibly containing viral integrations. As shown in Fig. 5, quantitative PCR analysis performed after removing viral replicative intermediate from the host genome by electrophoresis revealed that most of the viral sequences detected in repopulated livers at the end of the study were still present as replicative intermediates and not as integrated WM-HBV DNA. Artificial mixtures of uninfected genomic DNA and WM-HBV DNA isolated from serum were used to verify the efficacy of the gel extraction procedure in removing nonintegrated viral DNA forms from the high molecular weight host genome. Human genomic DNA obtained from an HBV-related liver tumor sample known to contain integrated HBV DNA sequences, or DNA isolated from PLC/PRF5 cell line (data not shown), were used as positive controls.

Figure 5.

WM-HBV DNA persisting in chimeric uPA livers after liver regeneration is mostly present as replicative intermediates and not as integrated virus. Relative amounts (%) of WM-HBV DNA sequences detected using real-time PCR in genomic DNA samples extracted from one PTH-chimeric liver of a mouse sacrificed 80 days after transplantation (Tx) of chronically infected PTHs. Episomal viral DNA was removed from viral integrations present within the high molecular weight host genomes using agarose gel electrophoresis. Artificial mixtures of serum-derived WM-HBV rcDNA and genomic DNA isolated from uninfected PTHs, as well as DNA from an HBV-related human hepatocellular carcinoma sample containing integrated HBV DNA sequences were used as negative and positive controls, respectively. Light gray bars show viral DNA amounts determined before gel extraction; dark grey bars show relative amounts of integrated viral DNA.

Discussion

The stability of the cccDNA pool is responsible for persistence of infection in most HBV-chronically infected individuals. Understanding molecular mechanisms affecting stability and activity of the cccDNA in vivo may reveal new therapeutic targets for the development of novel antiviral strategies able to enhance cccDNA clearance.11, 20 Although previous studies have noted that cccDNA turnover is slow even following administration of potent polymerase inhibitors,5-7, 12 little is known about the kinetics of cccDNA decay in the context of liver regeneration. WM-HBV-infected PTHs generated in uPA chimeric mice do not undergo hepatocellular changes, such as lipid droplet accumulation, nor do they show any sign of damage after long-term residence in immunosuppressive uPA mice, as has been reported for human hepatocytes.21 Hence, tupaia chimeric mice were chosen for serial transplantation of hepadnavirus-infected primary hepatocytes. Because of the lack of centromere structures able to ensure migration of the nuclear cccDNA minichromosome during cell division, a random and possibly unequal distribution of the cccDNA pool can be expected. However, this study showed that proliferation of infected hepatocytes led not only to strong dilution of the cccDNA pool per cell (Δ2 log), but it also induced significant loss of intrahepatic cccDNA loads (Δ0.5 log) within 80 days. Although the average doubling time of transplanted WM-HBV-positive PTHs in uPA mice was estimated to be 18 days, we found that liver regeneration significantly shortened the half-life of cccDNA per cell to 12 days, whereas the total loss of cccDNA in infected uPA livers was consistent with a half-life estimate of 38 days. Taking into account the short half-life of circulating hepadnaviruses previously determined in chronically infected HBV patients and in uPA/SCID mice16 and the relatively high amounts of intracellular rcDNA levels still detected during liver regeneration, it is striking that a relatively slow process of cell proliferation occurring in the absence of polymerase inhibitors was not able to assure replenishment of the cccDNA pool in serially transplanted PTHs. We expected that polymerase inhibitor-mediated inhibition of viral replication would have been required to avoid reestablishment of the cccDNA pool either by infecting virions or by re-entry of synthesized DNA-containing nucleocapsids into PTH nuclei.22 However, our data indicate that even if circulating virions may have re-infected proliferating PTHs and nuclear re-entry mechanisms may have occurred, four rounds of hepatocyte division induced a dramatic cccDNA decline with formation of significant amounts of cccDNA-free cells. Whether the addition of polymerase inhibitors would add a synergistic effect to the cccDNA reduction needs to be investigated.

Transcription of the different viral RNAs is regulated by independent promoters and previous studies performed on liver biopsy specimens obtained from treatment-naïve chronically infected HBV patients indicated that expression or stability of pgRNA and subgenomic preS/S viral RNAs may be differently regulated among individuals and in different phases of chronic infection.19 In this study, measurements of levels of subgenomic viral RNAs and serum HBsAg concentrations suggested that the production of envelope proteins was less affected by hepatocyte proliferation, whereas levels of intracellular rcDNA and pgRNA remained clearly below the levels determined before serial transplantation and routinely found in quiescent hepatocytes of both tupaia and human origin.17 Although reduced viral replication may have occurred at a transcriptional level, downstream replication steps such as RNA export and stability, encapsidation rates, and nucleocapsid maturation may also have been affected by the regenerative process and contributed to the lower virion productivity determined during liver regeneration. Further studies are needed to investigate whether stability of the cccDNA is similarly affected in HBV-infected human hepatocytes repopulating the uPA mouse livers and whether specific factors present in the milieu of liver regeneration can directly suppress viral replication in vivo.23, 24

We considered the possibility that part of the intrahepatic viral DNA detected toward the end of the study may have originated from integrations promoted in the course of hepatocyte turnover, as has been shown in cell culture and in the woodchuck system.9, 18, 25 However, viral integrations did not accumulate in chronically infected tupaia hepatocytes serially transplanted into uPA mice. Integration frequency varies among hepadnaviruses and in different hosts, and this may have contributed to the absence of integrations in proliferating PTHs. Furthermore, the nuclear entry process may be inefficient in regenerating hepatocytes so that despite the presence of low viremia levels in mice harboring regenerating WM-HBV-infected PTHs, establishment of new rounds of infection—and hence of DNA template necessary for formation of cccDNA and integrations—may have been restrained.

In previous models proposed to explain cccDNA clearance processes occurring during transient hepadnavirus infection, cytokine-mediated inhibition of viral replication,26 killing and proliferation of infected cells, as well as cccDNA capacities to survive mitosis, have been investigated.9, 10, 27, 28 In our experimental setting, virtually all PTHs engrafting the murine livers stained HBcAg-positive, and it is unlikely that expansion of PTHs occurred predominantly in uninfected cells, because in this case a significant intrahepatic cccDNA loss would not have occurred. Thus, in line with reports indicating that cccDNA loss is not only achieved through hepatocyte death, our findings provide the first direct evidence that regeneration of infected hepatocytes occurring without cell killing (Supporting Information) and in the absence of the adaptive immune response and polymerase inhibitors blocking new rounds of infection, induces drastic destabilization and significant clearance of the cccDNA pool in vivo. If mitosis of human hepatocytes represents a weak point in HBV persistence, these results suggest that therapeutic strategies aiming not only at reducing viral replication but also inducing a certain extent of cell injury and compensatory hepatocyte regeneration may be necessary to significantly reduce cccDNA loads in chronically infected individuals and possibly to achieve long-term immunological control of HBV infection.

Acknowledgements

We are thankful to Roswitha Reusch for her continuous excellent assistance with animal care.

Ancillary