Abstract

Methodologies for culturing muscle tissue are currently lacking in terms of quality and quantity of mature cells produced. We analyse images from in vitro experiments to quantify the effects of culture media composition on mouse-derived myoblast behaviour and myotube quality. Computational modelling was used to predict an optimum media composition for culturing. Metrics of early indicators of cell quality were defined. Images of muscle cell differentiation reveal that altering culture media significantly affects quality indicators and myoblast migratory behaviours. To predict cell quality from early-stage myoblast behaviours, metrics drawn from experimental images or inferred by Approximate Bayesian Computation were applied as inputs to an agent-based model (ABM) of differentiation with quality indicator metrics as outputs. We describe cell behaviours as a set of functions of media composition to predict cell quality using the ABM. Our results suggest that culturing muscle cells in neural cell differentiation medium reduces cell-cell fusion but does not diminish cell quality and that increasing serum concentration increases myoblast fusion implying a trade-off between the quantity and quality of cells produced when choosing a culture medium. Our model provided a good prediction of experimental results for media with 5% serum provided the myoblast proliferation rate was known. Author summaryFunctional skeletal muscle tissue can be grown in the lab but is most useful if the constituent muscle cells behave as they would in vivo. Optimising the conditions to promote precursor muscle cell fusion and growth is therefore vital. With many different factors influencing cell growth finding optimal conditions through rounds of experimentation alone is difficult especially as we strive to complexify our cultures with multiple cell types. We created metrics quantifying mature muscle cell quality at an early stage of development and applied them to experiments with a variety of culture media compositions. Changing the concentration of serum and proportion of neuron differentiation medium produced differences in the behaviour of fusing cells and in the quality and quantity of mature cells. From these results we created phenomenological models describing the behaviour of fusing cells for any combination of serum and neuron medium concentration. We integrated these models into a multiscale agent-based computational model of cell fusion to predict cell quality and quantity through virtual experiments in an emergent fashion. Our model suggests that choosing culture media composition will involve fundamental compromises between cell quantity and quality and that cells which initially fuse quickly may produce less final yield.