A multiscale model predicts the sensitivity of Chlorella vulgaris to light and nitrogen levels in photobioreactors

The maximization of lipid productivity in microalgae is crucial for the biofuel industry, and it can be achieved by manipulating their metabolism. However, little efforts have been made to apply metabolic models in a dynamic framework to predict possible outcomes to scenarios observed at an industrial scale. Here, we present a dynamic framework for the simulation of large-scale photobioreactors. The framework was generated by merging the genome-scale metabolic model of Chlorella vulgaris (iCZ843) with reactor-scale parameters, thus yielding a multiscale model. This multiscale model was employed to predict the sensitivity of growth and composition variation of C. vulgaris on light and nitrogen levels. Simulations of lipid accumulation quantified the tradeoff between growth and lipid biosynthesis under nitrogen limitation. Moreover, our modeling approach quantitatively predicted the dependence of microalgal metabolism on light intensity and circadian oscillations. Finally, we use the model to design a reactor irradiance profile that maximized lipid accumulation, thus achieving a lipid productivity increase of 46% at a constant intensity of 966 E m-2 s-1. Our modeling framework elucidated how metabolism and external factors can be combined to predict optimized parameters for industrial applications.