4.4.1 Suspension growth (w/o microcarriers) in a stirred tank bioreactor
In the case of stirred tank set-ups w/o microcarriers, Bayrak et al. [29] uses a hybrid methodology combining flux balance and agent-based modeling which can predict quantitative changes in nutrient and metabolite concentrations in a fed-batch set up. The model uses the measured dissolved oxygen and sodium data as input together with initial cell culture conditions and predicts viable cell density, viability and concentrations of glucose and lactate. The model shows good agreement with experimental trials. However, the application is low density (< 107 cells/mL) and, as the authors explain, assumes ideal mixing, with no allowance for spatial gradients. In cultivated meat, the output required demands high density and/or large-scale bioreactors, in which gradients are inevitable. To capture spatial gradients and heterogeneities will require three-dimensional modeling with cells or aggregates of cells represented by distinct agents.
The review paper by Hutmacher and Singh [30] on the application of computational fluid dynamics to three-dimensional modeling of tissue engineering-related bioreactors concludes that fluid flow processes have direct implications on cellular responses such as attachment, migration and proliferation. For cultivated meat, we propose to extend these principles to modeling of suspension growth in a stirred tank bioreactor, introducing a high-density of cell-clusters, each represented by an agent, to correspond to a higher density of cells.