Stochastic population balance modeling of influenza virus replication in vaccine production processes

A distributed population balance model of influenza A virus replication in adherent Madin-Darby canine kidney cells has been developed to reproduce and interpret flow cytometry data for virus propagation in microcarrier culture. The population of cells is differentiated into uninfected, infected and degraded cells. As an internal coordinate the number of intracellular viral components is considered. The main focus of the model is to link the time course of intracellular virus protein accumulation monitored by flow cytometry with the total yield of virus particles measured by the hemagglutination assay. The model allows simulating the extracellular virus dynamics for multiplicities of infection in the range 0.025-3.0. Shape of predicted histograms is in general agreement with distributions obtained by flow cytometry. Differences in time course at about 12-14 and 20 h post infection indicate that additional assumptions on intracellular virus dynamics are required to fully explain experimental data. Furthermore, prerequisites for virus replication, like receptor binding sites, the number of endosomes or the demand for free amino acids and nucleotides for virus synthesis can be estimated and compared with cellular resources available. Simulation results suggest that intracellular pools of free amino acids as well as early cell death due to influenza virus-induced apoptosis can limit virus yields. It. is expected that based on a better understanding of the infectivity status of cells and the spreading of viruses in population of cells in bioreactors strategies on design and optimization of vaccine production processes can be developed. (C) 2007 Elsevier Ltd. All rights reserved.