fax: (81) 82-255-6220.
Infection of human hepatocyte chimeric mouse with genetically engineered hepatitis B virus†
Article first published online: 25 OCT 2005
Copyright © 2005 American Association for the Study of Liver Diseases
Volume 42, Issue 5, pages 1046–1054, November 2005
How to Cite
Tsuge, M., Hiraga, N., Takaishi, H., Noguchi, C., Oga, H., Imamura, M., Takahashi, S., Iwao, E., Fujimoto, Y., Ochi, H., Chayama, K., Tateno, C. and Yoshizato, K. (2005), Infection of human hepatocyte chimeric mouse with genetically engineered hepatitis B virus. Hepatology, 42: 1046–1054. doi: 10.1002/hep.20892
Potential conflict of interest: Nothing to report.
- Issue published online: 25 OCT 2005
- Article first published online: 25 OCT 2005
- Manuscript Accepted: 14 AUG 2005
- Manuscript Received: 20 MAR 2005
- Japanese Ministry of Health, Labor and Welfare
Studies of hepatitis B virus (HBV) mutants have been hampered by the lack of a small animal model with long-term infection of cloned HBV. Using a mouse model in which liver cells were highly replaced with human hepatocytes that survived over a long time with mature human hepatocyte function, we performed transmission experiments of HBV. Human serum containing HBV and the virus produced in HepG2 cell lines that transiently or stably transfected with 1.4 genome length HBV DNA were inoculated. Genetically modified e-antigen–negative mutant strain also was produced and inoculated into the mouse model. A high-level (≈1010 copies/mL) viremia was observed in mice inoculated with HBV-positive human serum samples. The level of viremia tended to be high in mice with a continuously high human hepatocyte replacement index. High levels and long-lasting viremia also were observed in mice injected with the in vitro generated HBV. The viremia continued up to 22 weeks until death or killing. Passage experiments showed that the serum of these mice contained infectious HBV. Genetically engineered hepatitis B e antigen–negative mutant clone also was shown to be infectious. Lamivudine effectively reduced the level of viremia in these infected mice. In conclusion, this mouse model of HBV infection is a useful tool for the study of HBV virology and evaluation of anti-HBV drugs. Our results indicate that HBeAg is dispensable for active viral production and transmission. (HEPATOLOGY 2005;42:1046–1054.)
Hepatitis B virus (HBV) is a small enveloped DNA virus and causes chronic infection of the liver that often leads to chronic hepatitis, cirrhosis, and hepatocellular carcinoma.1–4 The lack of a practical small animal model has impeded the study of the biology of this virus and the development of effective antiviral therapies. Chimpanzee is the only natural host that allows active replication of HBV.5–7 Although this animal is a valuable model for the study of hepatitis viruses,8 the practical use of chimpanzees is severely limited both ethically and economically.
Several small animal models of HBV infection have been reported. The HBV transgenic mouse is a very useful model for the study of virology and evaluation of antiviral drugs.9–12 However, the liver cells of this model are not permissive for HBV infection; therefore, studying virus–cell interactions such as receptor binding and entry is not possible. The HBV-trimera mouse is another useful mouse model.13 In this model, ex vivo HBV-infected human liver fragments are implanted into lethally irradiated mice after SCID mouse bone marrow transplantation. Approximately 80% of the mice develop viremia 2 to 3 weeks after infection. However, the rate of positivity subsequently decreases to less than 20% 6 weeks after infection. The level viremia is approximately 105copies/mL. More recently, HBV-containing human serum samples were used to infect human hepatocyte repopulated mice.14 A high-level viremia (4.5 and 10 × 108copy/mL) and HBs antigenemia are observed 8 weeks after injection. This mouse model is promising because HBV replicates in natural host cells, human hepatocytes. However, long-term high-level viremia has not been reported so far in this model, probably because of technical difficulties in maintaining large quantities of human hepatocytes in these mice.
Long-term HBV viremia was reported after subcutaneous transplantation of immortalized human hepatocytes in RAG-2–deficient mice after transfection of circularized full-length HBV genome.15 Viremia of up to 3 × 108 copy/mL was still observed in these mice at least 5 months after transplantation. This long-term viremia model should be useful for in vivo HBV studies. However, the production and selection of HBV-secreting immortalized human hepatocytes takes a long time, and the level of viremia in the transplanted animal depends on the volume of live immortalized cells in mice. The mode of viremia might be different from natural infection because the pregenome RNA is transcribed from integrated HBV. Whether the produced HBV re-infects implanted immortalized human hepatocytes has not been confirmed.
A useful woodchuck hepatitis virus (WHV) infection model was established by Petersen et al.16 They showed high-level replacement of uPA/Rag-2 knockout mice liver with woodchuck hepatocytes and development of high-level (1 × 1011 virion/mL) WHV viremia. Dandri et al.17 transplanted Tupaia hepatocyte into uPA/RAG-2 mice and showed up to 8.2 × 107 genome equivalent/mL viremia. This model is useful because viremia continued up to 29 weeks. However, probably because of different host cells, the replication levels of HBV are lower than those of woolly monkey HBV. Using SCID mouse homozygous for Alb-uPA transgene, the group of Mercer and colleagues18 were the first group to report high-level replacement of mouse liver with human hepatocytes and successful infection of these mice with hepatitis C virus. Recently, we also created a human hepatocyte chimeric mouse in which the hepatocytes were highly replaced by implanted human liver cells.19 The repopulation index calculated from serum human serum albumin (HSA) concentrations exceeded 70% in 32% of the transplanted mice, and these animals survived up to 80 days after transplantation with high replacement index. Using this chimeric mouse, we performed transmission experiments of HBV. Using serum samples obtained from patients with chronic HBV infection, high-level viremia (approximately 1010 copies/mL) was observed up to 22 weeks in mice inoculated with HBV-positive human serum samples. We also performed infection study using in vitro–generated HBV. Infectious HBV was produced in HepG2 cell lines by transfecting with 1.4 genome length HBV DNA. Because mice injected with this in vitro–produced virus developed viremia, we further performed passage study. In addition, we introduced point mutations in HBV genome to create an HBe antigen–negative variant. The mice inoculated with this HBe antigen–negative variant developed viremia. Lamivudine effectively suppressed replication of HBV in mice inoculated with human serum samples and wild-type in vitro–created HBV. This model is a useful tool for the study of the nature of HBV mutants and development of anti-viral drugs.
Materials and Methods
Human Serum Samples.
Serum samples were obtained from four HBV carriers after obtaining written informed consent. Inocula for mice were obtained from two patients who tested positive for HBs and HBe antigens with slightly elevated levels of serum alanine aminotransferase and high-level viremia (Table 1). Serum samples for extraction and cloning of HBV were obtained from the remaining two patients who were positive for hepatitis B e antigen (HBeAg) and had high-level HBV DNA (6.9 × 109 and 9.8 × 1010 copies/mL by real-time polymerase chain reaction [PCR], respectively). All of these HBV belonged to genotype C.
|Inoculum||Source||Transfection||HBs Antigen||HBe Antigen||HBV DNA (LGE/mL)|
|Serum 1||HBV carrier 1||−||+||130||10.8|
|Serum 2||HBV carrier 2||−||+||150||8.7|
|CA59||pCAG-HB-wt||Stable||6.3 ± 2.8||15 ± 7||8.0 ± 0.2|
|CM3||pTRE-HB-wt||Transient||3.9 ± 1.5||105 ± 7||8.3 ± 0.4|
|e-Negative||pTRE-HB (PC)||Transient||2.9 ± 0.2||0.5 ± 0.3||8.1 ± 0.3|
|Mouse CM3||CM3-infected mouse||—||ND||ND||ND|
Analysis of HBV Markers.
Hepatitis B surface antigen (HBsAg) and HBeAg were measured by commercially available ELISA (Abbott Japan, Osaka, Japan). For quantitative analysis of HBV DNA, 100 μL serum samples or culture supernatants were used. DNA was extracted from these samples by SMITEST (Genome Science Laboratories, Tokyo, Japan) and was dissolved in 20 μL H2O. One microliter DNA solution was amplified by Light Cycler (Roche Diagnostics, Japan, Tokyo) for quantitation of HBV. The primers used for amplification were 5′-TTTGGGCATGGACATTGAC-3′ and 5′-GGTGAACAATGTTCCGGAGAC-3′. The amplification condition included initial denaturation at 95°C for 10 minutes, followed by 45 cycles of denaturation at 95°C for 15 seconds, annealing at 58°C for 5 seconds, extension at 72°C for 6 seconds. The lower detection limit of this assay is 300 copies. Nested PCR was used to detect a small amount of HBV DNA with the outer primers X1F1 (5′-CGCGGGACGTCCTTTGTCTA-3′) and X2R1 (5′-GTTCACGGTGGTCTCCATGC-3′) and inner primers X1F2 (5′-TACGTCCCGTCGGCGCTGAA-3′ and X2R2 (5′-CAGAGGTGAAGCGAAGTGCA-3′). The amplification condition included 35 cycles of 94°C for 1 minute, 58°C for 1 minute, and 72°C for 2 minutes after 2 minutes of initial denaturation at 94°C followed by 7 minutes of final extension using Gene Taq (Wako Pure Chemicals, Tokyo, Japan) with anti-Taq high (TOYOBO Co., Osaka, Japan) according to the instructions provided by TOYOBO.
Cloning of HBV DNA and Plasmid Construction.
Full-length HBV DNA was amplified using these HBV DNA samples by the method of Gunther et al.20 and cloned into pBluescript SK+ (Stratagene, La Jolla, CA). HBV DNA, 1.4 genome length, obtained from one of these two patients was cloned into pcDNA3 (Invitrogen, San Diego, CA) after replacement of CMV promoter with CAG to yield pCAG-HB-wt. Similarly, 1.4 genome length HBV DNA from the other patient was cloned into a plasmid vector pTRE2 (BD Biosciences, Franklin Lakes, NJ) and designated pTRE-HB-wt. A modified plasmid pTRE-HB-PC was generated by introducing a G-to-A point mutation to nucleotide 1896 to create precore stop codon (TTG to TAG). The substitution was introduced by a commercially available site directed mutagenesis kit (QuickChange Site-Directed Mutagenesis Kit, Stratagene). Nucleotide sequences of the HBV cloned into plasmids pCAG-HB-wt and pTRE-HB-wt were deposited into the GenBank database under accession numbers AB206817 and AB206816, respectively.
Transfection of HepG2 Cell Lines With 1.4 Genome Length HBV DNA and Endogenous Polymerase Reaction Analysis.
HepG2 cells were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% (v/v) fetal bovine serum at 37°C and under 5% CO2. The cells were seeded to semi-confluence in 6-well tissue culture plates. For transient transfection experiments, two plasmids; pTRE-HB-wt and pTRE-HB-PC, were used. Two micrograms of each plasmid was transfected using Fugene 6 transfection reagent (Roche Diagnostics, Indianapolis, IN) according to the instructions provided by the supplier. Three to five days after transfection, the culture supernatant was collected for infection of mice and quantitative analysis of HBV DNA by real-time PCR. Alternatively, calcium phosphate precipitation was performed to prepare fresh supernatant for large-dose administration experiments. Concentrated supernatants were prepared by using Microsep 10K spin filter, according to the instructions provided by the manufacturer (Pall Life Sciences., Ann Arbor, MI). The HBV particles produced in the supernatants were immunoprecipitated with protein A sepharose and mouse anti-HBs monoclonal antibody 2Z824Z (Institute of Immunology, Tokyo, Japan) and subjected to endogenous polymerase reaction21 and Southern blot analysis after sodium dodecyl sulfate/proteinase K digestion followed by phenol extraction and ethanol precipitation. The DNA was electrophoresed in a 1% agarose gel and transferred onto a nylon membrane. The transferred DNA was detected with full-length HBV DNA probe synthesized with the PCR DIG probe synthesis kit and the DIG Nucleic Acid Detection kit and CSPD, ready-to-use (Roche Diagnostics) in the Fluor-S Max MultiImager (BIO-RAD Laboratories, Hercules, CA).
For the production of stably transfected cell lines, HepG2 cells were cultured in DMEM supplemented with 10% fetal bovine serum at 37°C and 5% CO2. Cells were seeded into 90-mm-diameter culture dishes. Twenty micrograms of the plasmid pCAG-HB-wt was transfected by calcium precipitation. Twenty-four hours after transfection, the cells were split and cultured in G418 selection DMEM (1 mg/mL). One hundred fifty colonies were isolated and amplified for identification of virus-producing cell lines. Clones positive for both HBs and HBe antigens were selected and further analyzed for production of HBV particles. Finally, one of five cell lines that produced more than 105 copy/mL HBV DNA in supernatant were selected and used for further experiments. This cell line produced stable levels of HBV DNA for more than 12 months (data not shown).
Analysis of HBV Produced in the Supernatant of Transfected HepG2 Cell Lines by Sucrose Density Gradient.
Five milliliters HBV-positive serum (108 copy/mL) or 100 mL cell culture supernatant (107 copy/mL) was layered on a 20% (wt/wt) sucrose gradient, and centrifuged at 24,000 rpm for 1 hour at 4°C with a Beckman SW28 rotor Beckman Coulter, Fullerton, CA). The precipitate was resuspended with 500 μL phosphate-buffered saline. These HBV samples were layered on a linear 20% to 50% (wt/wt) sucrose gradient. Centrifugation was carried out at 24,000 rpm for 21 hours at 4°C with a Beckman SW40 rotor. The gradients were fractionated into 500-μL samples, and the density of each fraction was calculated from the weight and volume. Each fraction was diluted 10-fold and tested for HBV DNA by real-time PCR.
Generation of Human Hepatocyte Chimeric Mice and Analysis of Serum Samples.
Generation of the uPA+/+/SCID+/+ mice and transplantation of human hepatocytes were performed as described previously by our group.19 Animal protocols were performed in accordance with the guidelines of the local committee for animal experiments. Infection, extraction of serum samples, and sacrifice were performed under ether anesthesia. HSA was measured with a Human Albumin ELISA Quantitation kit (Bethyl Laboratories Inc., Montgomery, TX) according to the instructions provided by the manufacturer. Serum samples obtained from mice were aliquoted and stored in liquid nitrogen until use.
Histochemical Analysis of Mouse Liver.
The liver specimens of infected mice were fixed with 10% buffered-paraformaldehyde and embedded in paraffin blocks for histological examination. The liver sections were stained with hematoxylin-eosin or subjected to immunohistochemical staining by using an antibody against hepatitis B core antigen (HBc-Ag) (DAKO Diagnostika, Hamburg, Germany) or HSA (Bethyl Laboratories Inc.). Endogenous peroxidase activity was blocked with 0.3% H2O2 and methanol. Immunoreactive materials were visualized by using a streptavidin-biotin staining kit (Histofine SAB-PO kit; Nichirei, Tokyo) and diaminobenzidine.
Human Hepatocyte Chimeric Mice Develop High-Level and Long-Term Viremia After Inoculation of Serum Samples Obtained From Carriers.
Twenty chimeric mice were inoculated with 50 μL serum 1 (Table 1). We used mice that had relatively low-level HSA because we had previously found that mice with low-level replacement are susceptible to HBV (Chayama K and Tateno C, unpublished results). The HSA of these mice was 300,000 ng/mL (median, range, 40,000-3,090,000, Fig. 1A). All 20 mice tested positive for HBV DNA by nested PCR 2 to 4 weeks after inoculation. Eighteen of 20 mice developed quantitatively measurable viremia, but two mice showed very low-level viremia that was detectable only by nested PCR. Mice with persistently high-level HSA tended to show high virus titer (Fig. 1A-C). The maximum level of viremia was 9.5 × 1010 copy/mL. The viremia reached a plateau 4 to 6 weeks after infection. In contrast, mice with a rapid decrease in HSA or persistently low-level HSA showed low virus titer (Fig. 1D). We also performed infection experiments using serum 2 (Table 1). Of the five mice inoculated with this serum, all developed quantitatively measurable viremia 2 to 4 weeks after inoculation (Fig. 2). The level of viremia reached 1 × 107 to 1 × 109 copies/mL. The level of viremia also tended to be high in mice with high HSA levels.
HBV Generated in HepG2 Cell Lines Are Infectious to Human Hepatocyte Chimeric Mice.
HBV markers and endogenous polymerase experiments with Southern blot analysis of HBV produced by transiently or stably transfected HepG2 cell lines are shown in Table 1 and Fig. 3. The results indicated that these cell lines produced the expected HBV antigens and HBV DNA into the supernatant. Using virus particles produced by transient transfection of plasmid pTRE-HB-wt, we performed endogenous polymerase chain reaction experiments. Formation of fully double-stranded, relaxed circular DNA was observed after the reaction (Fig. 3). Sucrose density gradient analysis of HBV produced by stably transfected cell line (CA59, Table 1) showed that the produced viruses were sedimented to similar fractions of HBV obtained from the serum of the HBV carrier (Fig. 4), suggesting that HBV particles similar to those in serum are produced in these cell lines.
In the next step, we inoculated each chimeric mouse with 50 μL of the supernatants produced by transiently or stably transfected cell lines (Table 1). Three mice were inoculated with CM3 (Fig. 5A). Four weeks later, one of these three mice developed measurable viremia. It reached a high level (7.3 × 108 copy/mL) at week 14. A serum sample obtained from this mouse at week 6 was stored in liquid nitrogen and used in the subsequent passage experiments. The other two mice developed viremia, but its level was so low that HBV was only detectable by nested PCR. At week 13, these two mice (5-a-2, 5-a-3, Fig. 5A) were inoculated with serum 1, which induced high-level viremia in mice with high HSA levels (Fig. 1). These mice did not develop measurable viremia, suggesting that the low-level viremia in the latter two mice was due to low-levels of human hepatocyte replacement. Similarly, six mice were inoculated with supernatant CA59. One of these six mice developed quantitatively measurable viremia (peak, 2.6 × 109 copies/mL) (Fig. 5B). Two of the six mice developed viremia only detectable by nested PCR. For a more efficient reverse genetics infection procedure, we used mice with higher human albumin concentrations and inoculated each with 500 to 1,000 μL freshly prepared high-titer virus particles. This resulted in infection of all 10 mice (Fig. 5C). Thus, we established a highly effective infection procedure of reverse genetics of HBV.
Infection of Genetically Engineered Mutant Viruses.
Four mice were inoculated with the supernatant of genetically engineered e-antigen–negative HBV generated in a pTRE-HBV-PC-transfected HepG2 cell line (e-negative, Table 1). Three of these four mice developed quantitatively measurable, but relatively low-level (less than 107 copies/mL) viremia, 2 to 6 weeks after inoculation (Fig. 6). Nucleotide sequence analysis of the precore region showed that the sequence obtained from the infected mice was completely in agreement with the transfected plasmid with precore stop codon at 1896.
Passage Experiment of HBV From a Mouse Infected by In Vitro Generated HBV to Naïve Chimeric Mice.
Each of four naïve mice was injected with 5 μL serum samples obtained from a mouse that developed HBV viremia after inoculation of in vitro generated virus (CM3, Table 1). All four mice developed viremia at 2 to 6 weeks after inoculation (Fig. 7). One of the four mice that developed measurable viremia died at week 5 (7-a-3, Fig. 7). Another mouse (7-a-2) was weak and was sacrificed at week 13. The high-level viremia in the third mouse (7-a-1) increased further to 8.5 × 109 copies/mL. The remaining mice developed viremia detectable only by nested PCR (7-a-4).
Histochemical Analysis of the Liver of Mice Infected With HBV.
Liver specimens from mice that became positive for HBV DNA after the inoculation of the described passage experiment were subjected to histological and immunohistochemical analyses. Multiple foci of replaced human hepatocytes were noted in hematoxylin-eosin–stained sections (Fig. 8A) that were positive for HSA (Fig. 8B). Such positive human hepatocytes were also positive for the HBV core antigen in serial sections (Fig. 8C).
Effect of Lamivudine Treatment in Mice Infected With HBV.
Five mice that became positive for HBV DNA by inoculation with serum 2 (Fig. 2) were fed lamivudine (30 mg/kg/day)-containing food. A rapid reduction of HBV DNA level was observed in all 5 mice. Although two of five mice showed graft failure reflected by a decrease in HSA levels to the lower limits of the assay (2-a-2 and 2-a-5), the reduction of HBV DNA levels appeared before the decrease in HSA in these mice, suggesting that the decrease in HBV DNA was due to both the effect of lamivudine and the loss of virus replicating human hepatocytes (Fig. 2). Similarly, 1 mouse with high-level viremia as described in the above passage experiment (7-a-1, Fig. 7) showed a marked reduction of HBV DNA.
The major finding of the current study was the successful establishment of a model of HBV infection with long-term and high-level HBV viremia in the human hepatocyte chimeric mouse. The level of viremia correlated with the degree of human hepatocyte replacement indicated by HSA levels. We also showed that HBV created in vitro using HepG2 cell lines are infectious to this mouse model. Thus, a combination of chimeric mouse and molecularly cloned virus enabled us to prepare a practical model for the study of HBV virology. Chimpanzee is also a useful model for the study of HBV virology. Injection of molecularly cloned HBV into the liver of chimpanzee induced HBV infection and hepatitis.22 However, there might be some difference between hepatocytes of human and chimpanzee that could affect the nature of infection and the replication of this narrow host virus.
A critical difference between the chimeric mouse model reported here and chimpanzee is that there is no immune system active for HBV in the mouse model. Although the chimpanzee model is known to cause hepatitis and is suitable for the study of HBV-induced hepatitis,23 the mouse model is expected to be free from inflammation because these mice are SCID and do not have any human cytotoxic T lymphocytes. Actually, we observed no lymphocyte infiltration or focal necrosis of human hepatocytes in our mouse model. Recently, similar morphological changes in a similar model were reported by Meuleman et al.24; they also observed no alteration of liver architecture by HBV and hepatitis C virus infections. Interestingly, however, we observed a poor increase in the viral titer during the early phase of infection in some mice (Figs. 5A, 6). This might represent some innate anti-viral defense mechanism of liver cells themselves against viral infection. Further investigation is necessary to explore this issue.
A mouse model without any inflammation is an advantageous phenotype because it allows the study of HBV replication without any influence of immunological reaction. The model is also beneficial for studying the effects of drugs without any influence of fluctuation of the virus by immunological reaction. The HBV-infected mouse described here opens the way to create a long desired practical small animal model that overcomes economical and ethical problems associated with the chimpanzee model.
We showed in this study that reverse genetics of HBV can be achieved highly efficiently by using mice with high human albumin levels and inoculating mice with large amounts of freshly prepared virus particles (Fig. 5C). We further showed in this study that e-antigen is completely dispensable for infection and replication. The e-antigen–negative HBV-containing serum was previously used in chimpanzee and is known to induce more severe hepatitis.25 However, it is difficult to exclude the possible presence of a small amount of e-antigen–producing virus that might help infection and replication of HBe antigen-negative HBV strain. Our results clearly demonstrate that HBV can infect and replicate in the complete absence of e-antigen–producing species. However, the level of viremia was relatively low (less than 1 × 107 copies/mL) in these mice. Whether this is due to lack of e-antigen should be further confirmed in a larger number of mice with high replacement index.
Because the mice treated with lamivudine showed a reduction of viremia (Figs. 2, 7), our infected mouse is suitable for the study of new drugs. Lamivudine is a potent anti-HBV drug that reduces the virus and induces clinical remission and histological improvement.26–28 Emergence of drug-resistant HBV mutants against this drug as well as other anti-viral drugs is a serious problem in the treatment of HBV,29–31 as has been seen in the therapy of human immunodeficiency virus infection. Our model is especially useful for the study of the biology and drug susceptibility of such mutants, because almost all such drug resistances are based on only one or two point mutation(s).29, 31
In conclusion, the mouse model presented in this study is very useful for the study of HBV biology and evaluating of anti-HBV drugs. Furthermore, many applications of this model are expected because we can easily create, manipulate, and modify the model compared with other models.
The authors thank Eiko Okutani, Asako Yoshizato, Hiromi Ishino, Kana Kunihiro, and Kiyomi Toyota for their excellent technical assistance.
- 26Histological improvement in patients with chronic hepatitis B virus infection treated with lamivudine. Liver 1997; 2: 103–106., , , .
- 29Mutation in HBV RNA-dependent DNA polymerase confers resistance to lamivudine in vivo. HEPATOLOGY 1996; 3: 714–717., , , , , .