Hepatology International

, Volume 10, Issue 4, pp 559–566

HBV culture and infectious systems


  • C. Nelson Hayes
    • Department of Gastroenterology and Metabolism, Applied Life Sciences, Institute of Biomedical and Health SciencesHiroshima University
    • Liver Research Project CenterHiroshima University
    • Department of Gastroenterology and Metabolism, Applied Life Sciences, Institute of Biomedical and Health SciencesHiroshima University
    • Liver Research Project CenterHiroshima University
    • Laboratory for Digestive DiseasesCenter for Genomic Medicine, RIKEN
Review Article

DOI: 10.1007/s12072-016-9712-y

Cite this article as:
Hayes, C.N. & Chayama, K. Hepatol Int (2016) 10: 559. doi:10.1007/s12072-016-9712-y


While an effective vaccine against hepatitis B virus (HBV) has long been available, chronic HBV infection remains a severe global public health concern. Current treatment options have limited effectiveness, and long-term therapy is required to suppress HBV replication; however, complete elimination of the virus is rare. The lack of suitable animal models and infection systems has hindered efforts to unravel the HBV life cycle, particularly the early events in HBV entry, which appear to be highly species- and tissue-specific. Human primary hepatocytes remain the gold standard for HBV replication studies but are limited by availability and variability. While the HepaRG cell line is permissive for HBV replication, other hepatoma cell lines such as HepG2 do not support HBV replication. The recent discovery of sodium taurocholate transporting peptide (NTCP) as a primary receptor for HBV binding has led to the development of replication-competent cell lines such as HepG2–NTCP. Human hepatocytes grown in chimeric mice have provided another approach that allows primary human hepatocytes to be used while overcoming many of their limitations. Although the difficulty in developing HBV infection systems has hindered development of effective treatments, the variability and limited replication efficiency among cell lines point to additional liver-specific factors involved in HBV infection. It is hoped that HBV infection studies will lead to novel drug targets and therapeutic options for the treatment of chronic HBV infection.


Hepatitis B virusHepaRGHepG2NTCPHuman hepatocyte chimeric mice



Covalently closed circular DNA


Dimethyl sulfoxide


Hepatitis B virus


Sodium taurocholate transporting peptide


Chronic hepatitis B virus infection

Three hundred and sixty million people throughout the world are chronically infected with hepatitis B virus (HBV). While an effective vaccine has been available since 1986, and most adult patients with acute HBV are able to clear the infection successfully, those patients who develop chronic infection have limited treatment options. In these patients, the goal of treatment is to suppress HBV replication to reduce the risk of cirrhosis and hepatocellular carcinoma, but the treatment is not considered curative. Long-term treatment with nucleoside analogues and/or peg-interferon suppresses viral replication, but treatment discontinuation frequently leads to HBV reactivation. Therefore, new treatments are needed to successfully combat HBV. In order to accomplish this, experimental systems and models that more effectively emulate the environment of primary human hepatocytes in vivo are needed.

Current therapies act late in the HBV life cycle, but drugs that target viral entry prior to the formation of covalently closed circular DNA (cccDNA) might be effective in inhibiting vertical transmission and suppressing re-infection of the liver graft following transplantation or of healthy hepatocytes following immune therapy [1]. Entry inhibitors might also assist in treating chronic infection by preventing re-entry of capsids into the nucleus.


Hepatitis B virus

HBV is a partially double-stranded, enveloped DNA virus belonging to the Hepadnaviridae. With a diameter of 42 nm, HBV is one of the smallest known enveloped animal DNA viruses. HBV is highly species-specific, infecting only humans and some non-human primates [2], whereas related viruses target other primates or rodents (Orthohepadnaviruses) and birds (Avihepadnaviruses). While much has been learned using these related animal viruses, key host- or virus-specific differences in host receptors and cccDNA establishment require cautious interpretation of these results. HBV also varies among its eight genotypes (A–H), which differ by at least 8 %, and its multiple subgenotypes, which differ by at least 4 %. HBV is highly hepatotropic and replicates mainly or exclusively in hepatocytes. While this cell specificity suggests undiscovered tissue-specific host factors that might be required for infection, it also makes it more difficult to establish suitable infection systems in vitro.

HBV cell culture systems

The lack of suitable animal models and the difficulty of establishing cell culture infection systems have complicated efforts to elucidate details of HBV infection. A reproducible and robust in vitro virus infection model that supports the complete HBV life cycle is needed to resolve these mechanisms. Nonetheless, so far, in vitro virus spread of HBV, in which cultured cells support the complete sustained viral life cycle, is not well supported using existing infection models.

Primary human hepatocytes

HBV infection has previously only been observed in fresh primary hepatocytes from humans and chimpanzees [3, 4]. It was recently shown that primary hepatocytes from the tree shrew (Tupaia belangeri) can also be substituted for human hepatocytes. Cultures of primary human hepatocytes remain the gold standard for xenobiotic toxicity studies and have been shown to support HBV infection [59]. However, the effectiveness of the model is hampered by the limited availability and genetic diversity of human liver materials. Cultured primary hepatocytes fail to proliferate, and metabolic activity decreases over time [10]. In particular, liver-specific functions, including cytochrome P450 activity, decrease over time [11, 12]. The loss of hepatocyte-specific factors also results in variable phenotypic changes among cells. While problematic for toxicity studies, reduced expression of liver-specific factors also impairs replication of HBV, which relies on hepatocyte nuclear factors for transcription.

HepG2 cells

HepG2 cells are a human hepatocellular carcinoma cell line derived from a 15-year-old male with a well-differentiated carcinoma. Similar to differentiated liver cells, HepG2 cells secrete major plasma proteins such as albumin and transferrin as well as acute phase proteins such as fibrinogen, alpha 2-macroglobulin, alpha 1-antitrypsin, transferrin, and plasminogen [13]. Because HepG2 cells are well differentiated and form basolateral and apical domains corresponding to liver sinusoids and bile canaliculi, they are well suited as an in vitro system for polarized human hepatocytes. The cells support HBV virion production following transfection with HBV DNA [14, 15]. While some groups have reported binding and entry of HBV using HepG2 cells, virion production was not observed in these studies [1619], and reports of virion production in cells cultivated with dimethyl sulfoxide (DMSO) could not be confirmed [1922].

HepG2 cells differ morphologically from primary hepatocytes. While primary hepatocytes are cubic and often binucleate, HepG2 cells have an epithelial morphology with one nucleus containing 48–55 chromosomes [10, 23]. HepG2 cells also differ from primary hepatocytes in terms of gene expression. While mRNA levels and expression of housekeeping genes is similar between HepG2 and primary hepatocytes [10], expression of liver-specific transcription factors (hepatocyte nuclear factors and C/EBP-α) and phase I enzymes (e.g., CYP3A4) is greatly reduced in HepG2 cells [24, 25]. While DMSO helps to induce susceptibility in primary hepatocytes and HepaRG cells, it is not effective in HepG2 or other established hepatoma cell lines [19].

HepaRG cells

One of the few available cells models susceptible to HBV/HDV infection is HepaRG [2628], a human hepatoma cell line derived from a hepatocellular carcinoma from a female with chronic HCV infection [11]. When seeded at low density, HepaRG cells divide rapidly and show an elongated undifferentiated morphology. However, addition of DMSO and hydrocortisone hemisuccinate induces cells to differentiate into two cell types, one biliary epithelial-like cell type with flat morphology and clear cytoplasm that surrounds the second cell type that forms clusters of hepatocyte-like epithelial cells (50–55 % of cells). HepaRG cells have one or two nuclei with limited chromosomal alterations, including an additional chromosome 7 and a translocation affecting chromosomes 12 and 22 that include the loss of the 12p region [29]. Unlike HepG2 cells, HepaRG cells express cytochrome P450 and liver nuclear receptors at levels comparable to cultured primary hepatocytes, with relatively stable expression over 6 weeks. Removal of DMSO reduces P450 expression but does not affect expression of liver-specific transcription factors and transporters [30]. The cells have been shown to support replication of the HBV genome and secrete infectious HBV particles into culture medium [26, 27]. However, there are several drawbacks in using HepaRG cells for analyzing the HBV lifecycle and evaluating antiviral agents. Induction of differentiation with DMSO and hydrocortisone prior to infection is time consuming, and cells exhibit heterogeneity in albumin expression. The activity of a number of enzymes involved in drug metabolism also varies in HepaRG cells compared to primary human hepatocytes. In particular, expression of P450 (CYP2E1 and CYP2D6) was lower in HepaRG cells, whereas CYP3A4 and CYP7A1 are considerably elevated [30, 31].

The search for the HBV receptor

HBV encodes three envelope proteins, the small, middle, and large surface proteins, but only the large (L) S antigen (LHBsAg) harbors the preS1 domain, which contains a critical region between amino acids 2 and 47 that has long been known to be involved in hepatocyte binding [16]. Although this region was known to be important in HBV binding, the corresponding receptor molecule on the hepatocyte membrane has remained a mystery. Many candidate molecules have been proposed, including asialoglycoprotein receptor, transferrin receptor, and IL-6 receptor. Ferritin light chain (FTL), and squamous cell carcinoma antigen 1 (SCCA1) have also been proposed as co-receptors for HBV entry [32]. However, none of these molecules could be confirmed as the primary HBV receptor. Research based on duck HBV revealed an essential role for carboxypeptidase D (CPD) in avian hepadnavirus infection, but this mechanism does not appear to be shared in humans and highlights the need for cell lines that more closely approximate in vivo human hepatocytes.

Identification of NTCP as the primary HBV receptor

A milestone in HBV research occurred in 2012 when Yan et al. [33] reported a receptor candidate that facilitates entry of HBV into primary human and tree shrew hepatocytes. Near-zero-distance photo-cross-linking and tandem affinity purification were used to isolate a single protein, which was later identified by mass spectrometry as the sodium taurocholate cotransporting polypeptide (NTCP or SLC10A1) [33]. Their results were confirmed by Ni et al. [34], who tested HBV binding against hairpin RNA-silenced candidate targets that differed in gene expression between differentiated and undifferentiated HepaRG cells and showed that knockdown of NTCP prevented infection by HBV and HDV. Importantly, they also showed that HepG2 and Huh7 cells gained the ability to bind HBV when human but not mouse NTCP was expressed. To explain this species-specific difference, they determined that, while HBs contains two essential NTCP sequence motifs (aa 157–163, which is required for binding, and aa 84–87, which is required for HBV infection), mouse NTCP lacks the aa 84–87 motif [34, 35]. When key amino acids in this region of the mouse NTCP sequence were altered to match the human sequence, mouse NTCP gained the ability to support HBV infection [36]. These reports suggest that the difficulties in establishing HBV cell culture infection systems have primarily been due to insufficient NTCP expression or limited accessibility in candidate cell lines. Even in primary human hepatocytes, only freshly isolated cells are able to support HBV infection, which may be explained by the sharp decline in NTCP expression after culturing [37]. Even in cell lines that stably express NTCP, HBV infection may be limited by accessibility of NTCP. Disruption of the epithelial barrier of HepaRG cells increased the rate of HBV infection by allowing greater access to the basolateral membrane [38]. An overview of the role of NTCP and other host factors in the early stages of the HBV life cycle is shown in Fig. 1.
Fig. 1

Hepatitis B virus attachment, binding, and entry. HBV entry begins with reversible interaction with heparan sulfate proteoglycans and glypican 5 on the hepatocyte cell surface. Once attached, the virus binds with high affinity to the sodium taurocholate cotransporting polypeptide and enters the cell via endocytosis. The viral envelope fuses with the endosome membrane and releases the nucleocapsid into the cytoplasm. The nucleocapsid enters the nucleus through the nuclear pore complex, and the host cell’s DNA repair machinery converts HBV DNA from relaxed circular DNA (rcDNA) into covalently closed circular DNA (cccDNA)

Function of NTCP

Although its role in HBV infection was previously unknown, the function of NTCP itself has been relatively well characterized. NTCP is an integral membrane glycoprotein involved in sodium-dependent import of bile salts reclaimed by the small intestine [39, 40]. Because NTCP is expressed on the basolateral membrane, NTCP localization depends on proper cell polarization of the form observed in mature hepatocytes, which differs from that of other epithelial cells due to the reduced size and orientation of the apical membrane [38]. This partly explains the specificity of HBV infection and the rapid loss of hepatocyte infectability in cell culture [41]. Treatment with DMSO has long been known to improve HBV infection efficiency by promoting hepatocyte differentiation [22, 29], and treatment with DMSO has been shown to improve binding affinity of NTCP with preS1 [42]. It is not yet clear how NTCP facilitates HBV entry, but binding of either HBV or the preS1 mimic Myrcludex-B to NTCP inhibits NTCP function and disrupts bile import [34]. Because of its importance in maintaining intracellular bile salt levels, NTCP expression is highly regulated [40], and disruption of NTCP activity either by HBV infection or Myrcludex-B in chimeric mice resulted in compensatory changes in bile acid synthesis and cholesterol metabolism [43], including 12-fold up-regulation of cholesterol 7α-hydroxylase (CYP7A1), a key enzyme required in the conversion of cholesterol to bile acids. Given these results, it is possible that the cellular response to interference with NTCP import activity might hold clues into the poorly understood initial stages of HBV entry into hepatocytes following receptor binding.

NTCP-expressing cell lines

The discovery of the role of NTCP in HBV entry has spurred efforts to establish a robust in vitro infection system for HBV [36], and several cell lines have been successfully transfected with the human NTCP gene (e.g., hNTCP–HepaRG, hNTCP–Huh, hNTCP–HepG2, and hNTCP–HEK293) [33, 42].

NTCP-transfected HepG2 cells

Although unaware of the role NTCP in HBV infection, Jiang et al. [44] fused HepG2 cells with primary human hepatocytes to create the HepCHLine-4 cell line in an early effort to combine the HBV infection susceptibility of primary hepatocytes with the cell culture properties of HepG2. cccDNA was detected using polymerase chain reaction, and HBs antigen and viral particles were observed in the culture media. Interestingly, the cells were able to maintain susceptibility to HBV infection despite more than 12 months of sub-culturing.

HepG2–NTCP–C4 cells

Following the identification of NTCP as the HBV receptor, it became possible to engineer HepG2 cells to support HBV infection. HepG2–NTCP–C4 cells were shown to become susceptible to serum- and culture-derived HBV infection following transfection with NTCP expression plasmid [42]. Pre-treatment of cells with 3 % DMSO improved infection efficiency, resulting in infection of nearly half of the cells in culture. NTCP knockdown experiments and experiments involving NTCP disruption using anti-NTCP antibodies and cyclosporin A (and related molecules) inhibited HBV infection, supporting the role of NTCP as the principal mediator.

HepG2–NTCP12 cells

Using a similar approach, Yan et al. [45] developed the HepG2–NTCP12 line by transfecting HepG2 cells with the pcDNA6–NTCP plasmid. Although the infection rate was low, the authors found that infection efficiency increased following centrifugation (“spinoculation”) and suggest its potential usefulness as a standard protocol. Centrifugation did not affect the parental HepG2 lines, nor did it alter susceptibility to entry inhibitors.

Limitations of NTCP-expressing cell lines

Ectopic expression of NTCP in Huh7 and HepG2 cells alone induces HBV infection only at a low level and requires a high inoculum titer (e.g., 6,000–18,000 GEq/cell), whereas administration of DMSO drastically improved the infection rate [34]. This suggests that NTCP is not solely responsible for HBV entry and implies the presence of other differentiation-dependent host factors. While overexpression of NTCP improved infection efficiency in hepatoma cell lines [33, 46], HBs antigen and HBe antigen levels were low, and HBV DNA was not detected in the supernatant, suggesting that HBV production was low in these studies. In vivo studies involving the role of NTCP in HBV infection suggest that this receptor may be necessary but not sufficient for robust HBV infection, and other hepatocyte-specific factors in addition to NTCP may be required for HBV entry. A recent study by Verrier et al. [47] using RNA interference revealed that glypican 5 (GPC5) plays a role in hepatocyte entry for both HBV and HDV. GPC5 silencing reduced levels of HBsAg and HBV pgRNA and inhibited HBV binding to hepatocytes. Knockdown of GPC5 did not alter NTCP expression and had no effect on HCV replication. While NTCP may be the primary receptor, interactions with secondary receptors or factors might be required for maximum replication efficiency.

Human hepatocytes isolated from humanized mice

Despite the recent advances in HBV cell culture systems, it remains possible that immortalized cell lines are not able to fully recapitulate the properties of primary hepatocytes at the level required to draw unambiguous conclusions about the role of host factors and evaluate the efficacy of antiviral candidates with confidence. An alternative approach is to attempt to overcome the limitations of primary human hepatocytes in culture. In one approach, Ishida et al. [48] used the humanized mouse model as a source of fresh primary human hepatocytes, taking advantage of the high infectivity of primary hepatocytes while reducing variability in infection efficiency among cells and increasing the number of cells available. Humanized mice are generated by transplanting cryopreserved human hepatocytes from a single donor into urokinase-type plasminogen activator-transgenic/severely combined immunodeficient (uPA/SCID) mice [49] (Fig. 2). Hepatocytes replicate and repopulate the mouse liver while maintaining hepatocyte-specific differentiation, which can be monitored via human albumin levels. This mouse model has been successfully used for HBV and HCV infection studies [5052] and supports the complete HBV life cycle [5355], but the model also serves as a unique source of fresh primary human hepatocytes. Hepatocytes are isolated using a two-step collagenase perfusion method and cultured without passage in the presence of DMSO-supplemented hepatocyte clonal growth medium and polyethylene glycol. The extracted cells demonstrate high cytochrome P450 enzyme and glucuronosyltransferase activity. The cells were shown to support robust HBV infection, with up to 80 % of cells infected (in PEG-treated cells), and HBV DNA was persistently detected in the culture medium for each combination of donor and inoculum. Although the cost and complexity of this model is a limitation, proliferation of transplanted cells within the mouse is expected to yield a 500–1,000 increase in the number of available cells from the same donor. Because hepatocytes can be drawn from the same donor, this model improves homogeneity and availability against a reproducible genetic background with improved infection efficiency compared to directly cultured primary hepatocytes, HepaRG cells, or NTCP-expressing cell lines.
Fig. 2

The humanized mouse model as a source of primary human hepatocytes. Humanized mice are generated by transplanting cryopreserved human hepatocytes from a single donor into urokinase-type plasminogen activator-transgenic/severely combined immunodeficient (uPA/SCID) mice. Hepatocytes are then isolated and grown in culture [48]

Conclusion and perspective

Because chronic HBV infection increases the risk of cirrhosis and hepatocellular carcinoma, the large number of individuals currently infected with chronic HBV represents a long-term unresolved public health crisis. Treatment with nucleoside analogues and peg-interferon helps to slow progression of the disease but is not curative, and HBV entry inhibitors such as Myrcludex-B might help to prevent post-exposure infection or graft re-infection in liver transplant patients, but current drugs are not curative due to the long-term presence of cccDNA in infected cell nuclei. Development of new treatment approaches is likely to require a more thorough understanding of the viral life cycle, and particularly the role of host factors in HBV entry and the establishment of chronic infection. In vitro infection models will play a key role in development and evaluation of the next generation of antiviral compounds. However, existing HBV DNA transfected cell lines and immortalized cell lines in which gene expression varies substantially from that of primary hepatocytes might have limited usefulness in this endeavor. Instead, models demonstrating characteristic hepatocyte polarization and morphology, in which hepatocyte-specific transporters, enzymes, and nuclear factors are expressed at expected levels, are likely to yield more reproducible and verifiable results. The recent success in the development of direct-acting antiviral agents against hepatitis C virus provides hope and encouragement that further elucidation of the HBV life cycle will yield similar results.


This work was supported in part by Grants-in-Aid for scientific research and development from the Ministry of Education, Culture, Sports, Science and Technology, and the Ministry of Health, Labor and Welfare, Government of Japan.

Compliance with ethical requirements

This article does not contain any studies with human or animal subjects.

Conflict of interest

C.N. Hayes and K. Chayama declare that they have nothing to disclose regarding funding or conflict of interest with respect to this manuscript.

Copyright information

© Asian Pacific Association for the Study of the Liver 2016