Mathematical Models of Early Hepatitis B Virus Dynamics in Humanized Mice

Stanca M. Ciupe,Harel Dahari,Alexander Ploss

DOI: https://doi.org/10.1007/s11538-024-01284-2

2024-04-11

Bulletin of Mathematical Biology

Abstract:Analyzing the impact of the adaptive immune response during acute hepatitis B virus (HBV) infection is essential for understanding disease progression and control. Here we developed mathematical models of HBV infection which either lack terms for adaptive immune responses, or assume adaptive immune responses in the form of cytolytic immune killing, non-cytolytic immune cure, or non-cytolytic-mediated block of viral production. We validated the model that does not include immune responses against temporal serum hepatitis B DNA (sHBV) and temporal serum hepatitis B surface-antigen (HBsAg) experimental data from mice engrafted with human hepatocytes (HEP). Moreover, we validated the immune models against sHBV and HBsAg experimental data from mice engrafted with HEP and human immune system (HEP/HIS). As expected, the model that does not include adaptive immune responses matches the observed high sHBV and HBsAg concentrations in all HEP mice. By contrast, while all immune response models predict reduction in sHBV and HBsAg concentrations in HEP/HIS mice, the Akaike Information Criterion cannot discriminate between non-cytolytic cure (resulting in a class of cells refractory to reinfection) and antiviral block functions (of up to viral production 1–3 weeks following peak viral load). We can, however, reject cytolytic killing, as it can only match the sHBV and HBsAg data when we predict unrealistic levels of hepatocyte loss.

mathematical & computational biology,biology

What problem does this paper attempt to address?

The paper aims to study the early kinetic characteristics of Hepatitis B Virus (HBV) in humanized mice through mathematical modeling, with a particular focus on the role of adaptive immune responses during acute HBV infection. Several mathematical models were developed to describe HBV infection, either excluding adaptive immune response terms or assuming different forms of adaptive immune responses such as cytotoxic immune killing, non-cytotoxic immune cure, or non-cytotoxic mediated viral production blockade.

Researchers first validated a model without adaptive immune responses to match the time-series data of HBV DNA (sHBV) and surface antigen (HBsAg) in humanized mice (transplanted only with human hepatocytes, abbreviated as HEP). The results showed that the model lacking adaptive immune responses could well explain the high concentrations of sHBV and HBsAg observed in HEP mice.

Next, the researchers validated models that included different forms of adaptive immune responses to match the sHBV and HBsAg data in humanized mice transplanted with both human hepatocytes and human immune system components (HEP/HIS). The results indicated that all models incorporating adaptive immune responses could predict the decline in sHBV and HBsAg concentrations. However, according to the Akaike Information Criterion (AIC), it was not possible to distinguish between non-cytotoxic cure leading to reinfection resistance and antiviral action leading to reduced viral production. Nevertheless, the mechanism of cytotoxic killing could be excluded, as it could only match the experimental data when predicting unrealistic hepatocyte loss.

In summary, this study attempts to understand how adaptive immune responses influence the dynamic changes of the virus during acute HBV infection and tries to distinguish the effects of different immune response mechanisms on the HBV infection process through mathematical modeling.