Exploration of Hepatitis B Virus Infection Dynamics through an Intracellular Model

TL;DR

An intracellular HBV infection model is developed to capture RC DNA uncoating, cccDNA formation, transcription/translation, reverse transcription, and recycling, including SVP and virion production via compartments $R$, $C$, $R_g$, $R_s$, $R_h$, $H$, $P$, $Z$, $C_p$, $P_g$, $S_p$, $S$, $D$, $D_L$, $V$, and $S_v$, with a cccDNA activity fraction $\Phi$ linked to HBx. The system is solved with a fourth‑order Runge–Kutta method and calibrated against data from four entecavir‑treated HBV‑infected mice; global sensitivity analysis using Latin hypercube sampling–PRCC identifies HBx ($H$) and SVP/capsid pathways as major levers on infection dynamics. The analysis shows HBx markedly increases cccDNA and viral load, while intracellular delay ($\tau$) and dslDNA pathways have limited impact, and it highlights capsid recycling and SVP production as key drivers, suggesting therapeutic targets such as HBx inhibition and SVP release modulation. The work provides mechanistic insights into single‑cell HBV dynamics and informs antiviral strategies, while noting simplifications (linear intracellular processes, no immune components) and opportunities to extend to immune‑aware and multiscale liver models.

Abstract

Analysis of the cell population generally provides average information about viral infection in a host whereas the intracellular model captures the individual cellular responses. The primary goal of this study is to comprehensively analyze the intracellular dynamics of hepatitis B virus (HBV) infection and to identify the most influential factors. In this study, an intracellular HBV infection dynamics model is proposed by considering several intracellular steps that are observed in the virus life cycle. Upon comparison with the experimental data, it is seen that the model solutions exhibit a good agreement. The well-known fourth-order Runge-Kutta method is applied to numerically solve the proposed model. The effects of HBx proteins, dslDNA containing intermediates, intracellular delay and initial concentration of cccDNAs are explicitly studied. In order to identify the most positively and also the most negatively sensitive parameter of the proposed model, the global sensitivity analysis is performed using the widely-used method, Latin hypercube sampling-partial rank correlation coefficients. As a result, it is observed that HBx proteins have notable impacts on the dynamics of the infection, whereas intracellular delay and dslDNA-containing intermediates may not significantly affect the infection. This study also suggests that sub-viral particles could potentially contribute to the progression of the infection. Furthermore, recycling of capsids (an intracellular process perhaps unique to the HBV life cycle, where a portion of the newly produced capsids return to the nucleus and amplify the cccDNAs) is found to play an important role in enhancing the infection.

Figures (6)

• Figure 1: Calibration of the proposed model using experimental data collected from four humanized mice subjected to ETV treatment. Both model solutions and experimental data are presented on a logarithmic scale along the $y$-axis. The shaded areas represent the 95% confidence interval of the model solution.
• Figure 2: The effects of initial concentrations of cccDNAs on all components. Five different initial conditions $C(0)=10,20,30,40,50$ are considered here. Time and concentration of each component are taken along $x-$axis and along $y-$axis, respectively. The blue arrows in each subplot indicate direction of growth of the initial condition of cccDNAs.
• Figure 3: The effects of HBx proteins are demonstrated. Blue solid line and red dotted line represent the solution of the system \ref{['full model']} with and without considering the effects of the HBx proteins, respectively.
• Figure 4: Scatter plots of rcDNA, cccDNA, 2.4 and 2.1 kb mRNA, 0.7 kb mRNA, and HBx protein.
• Figure 5: Scatter plots of polymerase, core protein, RNP complex, pgRNA-containing capsid, surface protein, and ssDNA-containing capsid.