To investigate specific aspects of HBV replication, as well as the role of distinct viral proteins in HBV pathogenesis using convenient inbred animal models, embryo microinjection technologies have enabled the development of mice harboring either single HBV genes or terminally redundant over-length HBV-DNA constructs [116–120]. The first HBV-replicating transgenic mice, which were developed by Chisari and colleagues in 1995 [119], demonstrated the feasibility to produce in murine hepatocytes infectious HBV virions morphologically indistinguishable from human-derived virions [97]. As the immune system of transgenic animals recognizes during e mbryonic development the virus as “self,” these studies provided the first evidence that HBV replication does not induce hepatocellular injury [121]. To show that HBV-related pathogenesis is largely mediated by the host immune responses, induction of acute hepatitis and hepatocellular injury was demonstrated after adoptive transfer of HBV-antigen specific CTLs [122–124]. CTL-mediated release of cytokines also showed to suppress viral replication by non-cytolytic mechanisms [125, 126], while the recruitment of antigen nonspecific inflammatory cells amplified the severity of liver damage initiated by antigen-specific CTLs [127]. CD8+ T cells isolated from mice and engineered to express HBV specific chimeric antigen receptors (S-CARs) w ere also shown to engraft and expand in immunocompetent HBV-transgenic mice. After adoptive transfer, these cells were shown to control HBV replication, while causing only transient liver damage. Since these effector T cells, can be developed regardless of their HLA type, the adoptive transfer of such genetically modified HBV-specific T cells may represent a promising immunotherapeutic approach deserving further investigations [128, 129].

HBV transgenic mice have been successfully used to evaluate the impact of various polymerase inhibitors, such as lamivudine [130], adefovir dipivoxil and entecavir [131, 132] on HBV replication. Therapeutic approaches involving the use of HBV-specific small interfering RNAs (siRNAs ) were also tested in HBV-transgenic mice [133, 134]. To combine gene silencing with the induction of interferon responses, Protzer and colleagues recently employed 5′-triphosphorylated small interfering RNAs targeting highly conserved sequences on HBV RNA transcripts and showed that by triggering RIG-I-mediated innate immune responses these bifunctional antiviral molecules suppressed HBV replication more efficiently than siRNAs lacking a triphosphate group [135].

A major limitation encountered by using HBV transgenic mice is that they do not allow investigation of viral entry and spreading (Fig. 2.1). Moreover, no cccDNA is built in the liver of transgenic mice and the chromosomally integrated HBV genome cannot be eliminated from the host genome. As a consequence, viral clearance and cccDNA eradication cannot be achieved in this model. To break tolerance to HBV antigens and investigate the mechanisms of viral clearance, alternative mouse models have been developed which rely on transfecting or transducing the viral genome into mouse hepatocytes by different means, such as using recombinant adenoviral vectors , or by hydrodynamic injection of naked DNA in mice.

Adenoviral vectors containing a replication-competent HBV genome (Ad-HBV) have been shown to permit efficient transfer of the HBV genome into the liver of immunocompetent mice [136]. After intravenous injection of such adenoviral derived vectors, HBV proteins are produced under the control of endogenous HBV promoters and enhancers and viral replication could be demonstrated for up to 3 months. Viral clearance was accompanied by mild to moderate liver inflammation with elevated serum alanine transaminase activities [137]. After the induction of adaptive immune responses, anti-HBs seroconversion and development of neutralizing antibodies the infection was cleared. Inflammation and liver damage were also shown to be controlled by regulatory T-cells [138]. Because of the acute, self-limiting character of such adenoviral-mediated HBV infection, the system offers good possibilities to investigate the mechanisms of immune-mediated viral clearance also involving intrahepatic expansion of cytotoxic T cells (CTL) and NKT cells [139, 140]. By injecting relative low doses of adenoviral vectors [141] or by injecting the viruses intrahepatically into neonatal mice [142], persistent infection in immunocompetent mice could be established. This type of tolerance resembles immunological features of chronic HBV infection in humans. Moreover, even if mice do not establish cccDNA, using adenoviral vectors the HBV genome is in an extrachromosomal organization and hence its clearance can be achieved. Recombinant adeno-associated viruses (AAV) were also used to transfer replication-competent HBV genomes in a mouse strain carrying human leukocyte antigen A2/DR1 transgenes [143]. In these animals, viremia and antigenemia persisted for at least 1 year. Notably, a higher number of regulatory T-cells and no significant liver inflammation were determined in those livers, while impairment of functional T cell responses indicated the occurrence of tolerance. However, establishment of long-term viral replication with Ad-HBV vectors is limited by the immune responses against these vectors and occurrence of adenovirus-mediated cytotoxic effects may also limit their application. Since a functional cccDNA, associated with histone and non-histone proteins is not built in murine hepatocytes, these models are not suitable for the development of drugs or antiviral strategies targeting the natural template of HBV transcription and replication (Fig. 2.1).

Hydrodynamic injection techniques , which involve the rapid injection of a large volume (10 % of the animal weight) of a solution containing naked DNA into the tail vein of mice, are quite stressful for the animals but allow crossing species-specific barriers and permit efficient HBV DNA transfer [144]. Moreover, the rapid injection of liquid induces significant liver damage and ALT elevation shortly after injection. Hydrodynamic injection of replication-competent HBV genomes in mice resulted in viremia titres up to 1 × 10⁷ HBV DNA/ml [144]. Although HBV replication initiated already 1 day post-injection, replication levels decreased after 1 week and HBV was cleared from blood within 2–3 weeks, as soon as specific antiviral antibodies and CD8+ T cells appeared. However, HBV infection persisted for 3 months after hydrodynamic injection of mice lacking adaptive immune cells and natural killer cells, thus demonstrating that the outcome of hydrodynamic transfection of HBV depended on the host immune response [144]. Hydrodynamic injection studies also showed that simultaneous delivery of HBV expressing plasmids and HBV-specific siRNAs prevented HBV replication [133, 134], while by injecting modified HBV DNA plasmids into C57BL/6 mice, a significant immune clearance of HBV could be achieved [145]. A recent report showed that the use of a lentiviral backbone instead of an AAV vector led to increased and prolonged HBV replication (>56 days post injection) [146]. In an attempt to generate cccDNA-like molecules in mice, a monomeric HBV genome precursor plasmid (pr-cccDNA), that can be converted by Cre/loxP-mediated DNA recombination into a 3.3-kb cccDNA, has been recently used [147]. Although such recombinant cccDNA could be detected in the nuclei of murine hepatocytes, the induced immune response rapidly limited viral replication in vivo [147].

To investigate whether cccDNA molecules could be directly targeted for destruction, the CRISPR/Cas9 system has recently been employed both for in vitro [148, 149] and in vivo studies, using the HBV-hydrodynamic-mouse models [150]. In these studies, the levels of HBV-expressing vectors and different markers of viral replication were significantly reduced, and without evidence of toxicity, suggesting that the CRISPR/Cas9 system could be recruited to the HBV-expressing vectors. Although many issues regarding the efficiency and safety of the system remain to be addressed, these findings are the first to demonstrate nuclease-mediated disruption of a HBV expressing vector, as a model of cccDNA, thus opening new possibilities for the development of innovative antiviral strategies aimed at disrupting the cccDNA.

Despite the relatively short span of viral replication available, mice transfected by hydrodynamic injection are suitable not only for short-term antiviral studies but also for testing the consequences of specific mutations within the viral genomes. In comparison to HBV-transgenic mice, hydrodynamic-based procedures permit investigation of immune response emerging during acute infection. Since the viral genome is not integrated into the host genome as a transgene, viral clearance commonly occurs in these systems. A clear advantage of the system is that different HBV genotypes and variants can be injected into mice and analyzed in vivo in relatively short time (Fig. 2.1). Nevertheless, the rapid injection of large volume of fluids causes not only strong discomfort to the animals but also significant damage in the liver, which may alter cell function and signaling analyses. Since reinfection of the mouse hepatocytes is not possible, viral clearance is easier to achieve in these non-transgenic murine models than in humans, and therefore, the efficacy of antiviral treatments should be validated in systems permissive for HBV infection. Interestingly, all results available so far indicate that murine cells engineered to express the human NTCP do not become susceptible for HBV infection [44]. It appears that species-specific differences or the lack of cellular factors involved in post-entry steps are responsible for these discrepancies.

These mice represented the first human–mouse chimeric system that was developed by transplanting human liver fragments under the kidney capsule of normal Balb/c mice. To avoid rejection of the implanted tissue, these mice were preconditioned by total body irradiation and reconstituted with SCID mouse bone marrow cells [151]. After ex vivo infection of the small human liver specimens with HBV, low levels of viremia, that remained detectable for approximately 20 days, could be determined in implanted animals. Interestingly, mice ectopically carrying human liver fragments could also be engrafted with human peripheral blood mononuclear cells (BPMC), so that the effects of polyclonal anti-HBs antibodies against HBV could be assessed [89, 152]. Nevertheless, due to the extra-hepatic location of the implanted tissues, human hepatocytes remained functional only for limited time and in vivo infection with HBV or other human hepatotropic viruses could not be established.

To achieve long-term survival of primary human hepatocytes permissive for HBV infection in vivo, isolated cells need to be integrated in the mouse liver. The requirements to achieve this goal are (1) the damage of the endogenous murine hepatocytes to create the space and the regenerative stimulus necessary for transplanted hepatocytes to reconstitute the diseased mouse liver; and (2) the absence of murine adaptive immune cells and NK cells to permit engraftment and survival of transplanted xenogeneic hepatocytes. Different human-liver chimeric mouse models are currently available. The Alb-urokinase-type plasminogen activator (uPA) transgenic mouse was the first model describing the strong regeneration capacity of healthy transplanted hepatocytes. In this system, over-expression of the hepatotoxic uPA transgene, which is driven by the mouse albumin promoter, induces high levels of uPA in plasma, hypo-fibrinogenemia and subacute liver failure in young mice [153]. To generate mice with human liver chimerism, uPA transgenic mice have been backcrossed with immunodeficient mouse strains, such as the RAG2^−/− [154–156], the Severe Combined Immune Deficient (SCID) [157], which lack functional B and T cells or SCID/beige mice, which also lack NK-cell functions (shortly USB mice ) [158, 159] and NOD/SCID/gamma(c)(null) (shortly uPA-NOG) [178]. Following intra-splenic injection of one million freshly isolated or cryopreserved-thawed human hepatocytes, the transplanted cells migrate via the splenic and portal veins to the liver, where cells integrate into the liver parenchyma. Engrafted human hepatocytes proliferate to form larger regenerative nodules that eventually merge together to replace the diseased liver parenchyma. Reconstitution of the mouse liver takes around 2 months and the levels of human chimerism can be estimated by determining the concentration of human serum albumin in mouse blood [159, 160]. Within the mouse liver, the transplanted human hepatocytes maintain normal metabolic functions [157, 161]. To delay the production of the toxic transgene, which makes transplantation procedures necessary in the first month of life, alternative mouse models, where the expression of the uPA transgene is inducible [162] or is regulated by the MUP (major urinary protein) promoter, have been also developed [163]. To generate an animal model where hepatocyte failure can be induced at will in adult mice, alternative human-liver chimeric mice, based on the use of fumaryl acetoacetate hydrolase-deficient (FAH) mice, were established [164, 165]. FAH plays a crucial role in the tyrosine breakdown pathway and its deficiency leads to accumulation of toxic tyrosine catabolites and liver failure. However, accumulation of these catabolites can be avoided by administering the drug NTBC (2-(2-nitro-4-trifluoromethylbenzyol)-cyclohexane-1,3-dione), a pathway inhibitor that protects the animals from the occurrence of liver injury until drug administration is withdrawn. More recently, by transplanting higher amounts of human hepatocytes, or by performing repeated hepatocyte transplantation , high rates of human hepatocyte chimerism could be achieved in mice also lacking the Rag2 and the gamma-chain of the receptor for IL-2 genes (shortly FRG mice ) [166]. Human hepatocytes were also successfully transplanted in mice expressing the herpes simplex type-1 thymidine kinase (TK) transgene that were backcrossed with NOG (NOD/SCID/gamma(c)(null)) mice [167]. Even in this case, human hepatocyte transplantation can be performed in adult mice, since mouse liver cells expressing the TK-transgene can be selectively destroyed upon administration of ganciclovir [168].

From the first successful transplantation of human hepatocytes into uPA/RAG2 mice and establishment of de novo infection with HBV [156], several groups performed HBV infection studies [160, 169, 170], as well as demonstrated infection with other human hepatitis viruses , like HCV and HDV, in humanized uPA/SCID or uPA/SCID/beige (USB) mice [160, 171–173]. Notably, after intra-peritoneal inoculation of HBV infectious serum or cell culture derived virions, productive HBV infection, which requires the establishment of a functional cccDNA in hepatocyte nuclei, is first achieved in a minority of human hepatocytes and several weeks are needed to accomplish viral spreading [159]. After that, nearly all human hepatocytes stain HBcAg-positive and viremia reaches a stable plateau which, to a certain extent, correlates with the levels of human chimerism. The use of patient serum samples as virus inoculum allows the functional analysis of distinct HBV genotypes, naturally occurring variants and drug-resistance mutants. Genetically engineered viruses can therefore be used to investigate the role of distinct viral proteins in human infected hepatocytes. Using this type of approach, the expression of the regulatory HBx protein provided in trans was shown to be essential for cccDNA-driven HBV replication in infected human hepatocytes [170]. Studies focusing on investigating how HBV may affect cellular pathways [161] and innate immune responses of the human hepatocytes are just starting to emerge and showed that expression of metabolic genes [161] and innate immune response genes in these cells resembles well the expression pattern determined in human livers [158]. To this regard, a recent study indicated that binding of HBV to its cellular receptor alters the hepatocellular uptake of bile salts and the expression profile of genes of the bile acid metabolism [161]. The occurrence of such alterations was also confirmed in patient biopsy samples. Using humanized mice , both HBV and HCV have been recently shown to contribute to the induction and accumulation of aberrant DNA methylation in human hepatocytes through the activation of NK-cell-dependent innate immune responses [174].

Both the antiviral activity of clinically approved polymerase inhibitors [115, 158, 169] and the in vivo efficacy of novel polymerase or capsid inhibitors [175, 176] have been assessed in human liver chimeric mice. The model also served to evaluate the in vivo efficacy of lipopeptides derived from the large envelope protein of HBV to prevent de novo HBV [177] and HDV infection [172], as well as to investigate the ability of the most clinically advanced entry inhibitor, Myrcludex-B, to block HBV spreading post-infection [178]. Moreover, the serial passage of infected hepatocytes isolated from infected mice and transplanted into new recipients has permitted to gain insight about the impact of hepatocyte proliferation on cccDNA stability and activity in vivo [179]. The drastic reduction of intrahepatic cccDNA loads induced by cell division even in the absence of antiviral therapy revealed a weak point in HBV persistence that deserves further investigations. Thus, these systems offer unique opportunities to investigate factors that can affect the stability and/or activity of the cccDNA minichromosome, as well as the direct antiviral effects of cytokines and immune modulatory substances, such as interferons (Fig. 2.1). To this regard, human liver chimeric mice have been used to assess whether HBV can circumvent the induction of antiviral defense mechanisms [158]. Upon administration of regular human IFNα, HBV was shown to hinder the nuclear accumulation of STAT-1, thus providing a potential mechanism for the reduced induction of interferon stimulated genes (ISGs) determined in HBV-infected human hepatocytes [158]. On the other hand, studies in these mice also showed that IFNα can mediate epigenetic repression of the cccDNA minichromosome [29], while the repeated administrations of the longer-active pegylated IFNα was shown to be able to breach the impairment of HBV-infected hepatocyte responsiveness and induce sustained enhancement of human interferon stimulated genes (ISG) [180]. In this study, the stronger antiviral effects of peg-IFNα exerted on the human hepatocytes were shown to trigger a substantial decline of circulating HBsAg and HBeAg levels in chimeric mice, without claiming the involvement of immune cell responses [180]. Moreover, the comparative analyses of the innate immune responses revealed that type I, II and III IFNs are differently induced in murine and human hepatocytes and that the effects of distinct IFNs may differ between animal species, thus underlying the importance of validating results obtained in murine systems also by employing primary human hepatocytes [181].

Human hepatocytes can be very abundant within the chimeric livers but non-parenchymal cells, such as sinusoidal endothelial cells and Kupffer cells, are of murine origin. As a consequence, development of fibrosis and pathophysiologic processes that are commonly associated with chronic viral hepatitis infections but involve a cross talk between the hepatocytes and other liver resident cells are not observed in the above mentioned systems. Moreover, since these chimeric animals are genetically immune deficient they are not suited for vaccine studies and for evaluation of adaptive immune responses. To circumvent these limitations, partially haplotype-identical human peripheral blood mononuclear cells (PBMC) have been recently transferred in uPA/SCID chimeric mice after the establishment of HBV infection [182]. Notably, infiltrating human immune cells caused severe hepatocyte degeneration, while treatment with anti-Fas antibodies or depletion of dendritic cells prevented the death of human hepatocytes.

A different attempt to generate a humanized mouse model harboring both human liver cells and a human immune system was the development of the AFC8 model [183]. These mice were obtained by crossing BALB/c-RAG2^−/−γc^−/− mice, which lack functional B, T and natural Killer cells, with mice carrying a liver-specific suicidal transgene with inducible activity based on the induction of caspase 8 in mouse hepatocytes. These animals were used to transplant simultaneously human fetal hepatocytes and hematopoietic stem cells that were obtained by the digestion of human fetal liver tissues (15–18 weeks of gestation period). Since the fetal liver provides both types of cells, reconstitution of both cell lineages is syngeneic and hepatocyte rejection by the human immune system is not expected. Although HBV infection studies were not performed, after inoculation of hepatitis C virus, low levels of intrahepatic viral replication, as well as T-cell responses and development of fibrosis could be determined. Despite partial success and the ethical restrains encountered by employing human fetal tissues, the generation of chimeric systems equipped with both a human liver and a functional human immune system, with a matched major histocompatibility complex, still remains a major challenge. The latest advances reported by Gutti et al. showed the feasibility to reconstitute an uPA-NOG mouse strain with both adult human hepatocytes and hematolymphoid cells [184]. In this study, dual reconstitution was achieved by transplanting fetal or even adult hepatocytes and mismatched hematopoietic stem cells (CD34+ HSCs) derived from either a fetal liver or umbilical cord blood. As an alternative and well-tolerated procedure to total body irradiation, mice received 3 days of a treosulfan-based chemotherapeutic conditioning before HSC injection. The presence of CD8+ and CD4+ human cells and of human hepatocyte clusters was observed in the liver of these animals. No cell-mediated rejection but also no evidence of cell interactions were determined in animals reconstituted with mismatched HSCs. Although the applicability of these approaches for the study of HBV and HDV associated immune pathological processes needs further research, these technical advances have paved the way for the development of dually reconstituted humanized systems.