Experimental chronic hepatitis B infection of neonatal tree shrews (Tupaia belangeri chinensis): A model to study molecular causes for susceptibility and disease progression to chronic hepatitis in humans

Chronic hepatitis B virus (HBV) infection is an important global health problem, and is endemic in Southeast Asia and Africa. Although effective vaccinations are available, it has proven to be impossible providing large segments of the population in endemic regions with it. Therefore, the infected population is steadily expanding. As a result, many chronically HBV-infected patients will progress to developing cirrhosis and hepatocellular carcinoma [1]. For discovery and development of novel therapies, it is pivotal to have suitable in vivo disease models. A significant obstacle in developing such animal models is the fact that susceptibility to HBV is rather specific to humans. Only very closely related model is that of non-human primates, such as chimpanzees, who are also susceptible. However, use of these primates on a larger scale would be cost prohibitive.

Various hepadnavirus are infectious to small animals such as woodchuck, ground squirrel, duck and others, but these viruses vary significantly from HBV in structure and properties. The tree shrew (Tupaia) is evolutionary closely related to humans [2, 3]. Blum and his colleagues have successfully infected HBV in extracorporeal hepatocytes of tree shrews, and proved permissiveness and replication of HBV in hepatocytes of this small animal [4‐6]. However, thus far, experimental data suggested that tree shrew in vivo can only be transiently infected with HBV, which would limit the capabilities of this model in regards of studying chronic hepatitis in humans. Similarly, the majority of acute infected humans will ultimately clear the infection, with only a smaller portion developing chronic hepatitis. However, it is exactly this population that also carries the highest risk for developing cirrhosis and liver cancer. Therefore, suitable and more economic models mimicking chronic human HBV infection would be of significant value. It is not fully understood what are the exact risk factors which ultimately determine failure of virus clearance after the acute infection, and what factors exactly trigger and support progression to a chronic infection.

Nevertheless, clinical observations show that newborns and infants are more susceptible to infection with HBV, as well as progression to chronic infection. Similarly, susceptibility of animals to hepadnavirus is also inversely age-related [7‐9]. It has been speculated that this might be due to the immaturity of the immune system. Previously, we published preliminary data from a small pilot study in support of this speculation [10]. In order to further consolidate these data and to generate unequivocal scientific proof for this hypothesis, we continued the experiments with expanded animal cohorts. We investigated susceptibility and long-term progression of HBV-inoculated outbred, genetically heterogenous newborn tree shrews over a protracted time period up to 4 years.

In this study we provide evidence that 13-21% of outbred, genetically heterogenous newborn tree shrews are susceptible to chronic HBV infection lasting at least 48 weeks (up to 228 weeks). Given a life expectancy for tree shrews of about 6–8 years, we arbitrarily set the time duration of 48 weeks as criterion defining chronic HBV-infection. The data presented demonstrate that HBV is capable of sustaining itself at a viable and replication competent infectious state in tree shrews with formation of intermediate replication products over prolonged periods.

Development of chronic hepatitis in tree shrews demonstrates other epidemiologic similarities such as age-related susceptibility to infection with its human counterpart. We failed to inoculate adult tree shrews (data not shown), but similar to human neonates who also are significantly more susceptible to HBV infection, we achieved persistent infection in a smaller portion of neonate animals.

In contrast to a genetically homogeneous animal strain, an outbred population allows to investigate genetic and immunologic heterogenic predispositions for chronic HBV infection susceptibility, disease establishment and progression. Therefore, the described model would be an economic research tool which could facilitate identification of individuals at risk for infection and also aid in the development of new therapies. However, this would also require full understanding of tree shrew biology and generation of standardized research reagents. In this context, full sequencing of the tree shrew genome and allocation of appropriate resources might be considered by governmental or private agencies.

A direct relationship between intrahepatic HBV DNA and clinical outcome parameters had been previously reported [11‐13]. Although the copy numbers of serum HBV DNA in our study were generally lower when compared to humans suffering from chronic HBV-infection, we consistently confirmed presence of hepatic HBV DNA prior to its detection in serum in animals later to be confirmed as chronically infected. Namely, HBV copy number was higher in livers but lower in sera of tree shrews comparing to the values from human control samples. These data indicate that the liver is the main replication site of HBV in tree shrews, which precedes the appearance of detectable virus in serum. Therefore, we view the examination of liver biopsies as an important tool in order to predicting potential establishment of chronic HBV infection in tree shrews. Interestingly, we identified a significant number of tree shrews that maintained a persistent low burden of liver HBV DNA (10³-10⁴ copies/μg liver DNA) and negative HBsAg. This presentation closely resembles a clinical condition known as “occult HBV infection” in humans. Because occult HBV infection carries a high risk of developing chronic liver diseases including cancer in humans [14, 15], this model could provide a unique opportunity studying this condition further in vivo.

Histopathological changes were relatively mild, even in liver biopsies from confirmed long-term HBV-infected tree shrews. Possible explanations might be the relatively short observation period, although it still constituted a significant portion of the entire life span of a tree shrew. Moreover, the obtainable tissue specimens were extremely small and thus accurate assessment of histopathologically observed changes is rather limited and not entirely transferable to the entire organ. Interestingly, a similar situation (lack of chronic liver histopathologic changes compared to human pathology) is also seen within other animal models of hepatitis virus infection, such as the woodchuck model of woodchuck hepatitis B virus (WHBV) infection, and even the non-human primate chimpanzees models of HBV infection [16]. The reasons for the disparity of histopathological changes between humans and animals chronically infected with hepatitis virus remains unknown.

Regarding the factors affecting HBV infectivity, some points should be considered. For example, when using human HBV-infected serum to inoculate primary tupaia hepatocytes (PTH), Cock et al. [5] demonstrated that serum purification increased infectivity and this was explained with the supposition that human serum could inhibit binding of HBV virions to PTH. Meanwhile, various studies indicate that high titer of HBV DNA and/or certain HBV genotypes may associated with an increased risk of developing chronic liver disease in humans [17‐19]. Cote et al. [7] showed that besides age, the dose of virus inoculum to be a major factor influencing the course of chronic infection in woodchucks. In our study, the rate of chronic-infection induced by sera derived from HBV-infected tupaia appears to be higher than that by sera derived from HBV-infected humans, although the sample number is too small to make reliable statements concerning the significance of these differences. However, when comparing tree shrews inoculated with sera from HBV-infected humans to tree shrews inoculated with sera from HBV-infected tree shrews, the latter usually displayed lower levels of HBsAb, or their HBsAb appeared only intermittently. As an immunologic marker, appearance of HBsAb indicates clinical resolution of the HBV infection. In this context, this observation could suggest that using sera from the same infected species for inoculation might cause a lesser immune response than using infected serum from another species for that purpose. Thus, using tupaia-derived HBV-infected serum for inoculation might improve the infection rate. This hypothesis is currently being further investigated in our laboratory.

Although the frequency of infections in our model might resemble certain real life scenarios, higher infection rates might be desirable for certain experimental settings. To accomplish this, we are currently investigating various approaches such as using HBV strains with higher pathogenicity for inoculation, purifying or concentrating the inoculation serum in order to achieve higher viral titer or infection efficiency.

In summary, our data demonstrate that the tree shrew model has promise in studying pathophysiology of chronic HBV-infection, thus justifies further investments into studying the genome and biology of tree shrews in an effort to further consolidate it as a model for human disease.