A continuous dynamic genome-scale model explains batch fermentations led by species of the Saccharomyces genus.

Abstract

Batch fermentation is a biotechnological dynamic process that produces various products by employing microorganisms that undergo different growth phases: lag, exponential, growth-non-growth, stationary, and decay. Genome-scale constrained-based models are commonly used to explore the phenotypic potential of these microorganisms. Previous studies have primarily used dynamic Flux Balance Analysis (dFBA) to elucidate the metabolism during the exponential phase. However, this approach falls short in addressing the multi-phase nature of the process and secondary metabolism, posing significant challenges to our understanding of batch fermentation. A recent attempt at a solution was a discontinuous, multi-phase, multi-objective dFBA implementation. However, this approximation lacks the mechanistic connection between phases, limiting its applicability in predicting intracellular fluxes during batch fermentation. To overcome these limitations, we combined a novel continuous model with a genome-scale model to predict the distribution of intracellular fluxes throughout the batch fermentation process. The proposed model includes empirical descriptions of regulation that automatically identify the transition between phases. Its application to explain primary and secondary metabolism of Saccharomyces species in batch fermentation results in biological insights that are in good agreement with the previous literature. The ability to account for all process phases and explain secondary metabolism makes this model a valuable and easy-to-use tool for exploring novel fermentation processes. IMPORTANCEThis research proposes a novel dynamic genome-scale modelling approach for batch fermentation, a crucial process widely used to produce a diverse range of products such as biofuels, enzymes, pharmaceuticals, and food products or ingredients. The proposed approach automatically accounts for the transitions between different phases of the fermentation process (lag, exponential, growth-no-growth, and stationary). This is a significant advancement over previous methods that required different model formulations for different phases. We have successfully applied this modelling approach to explore the primary and secondary metabolism of three yeast species under batch fermentation conditions. The model accurately explained experimental data and provided biological insights consistent with previous research findings, instilling confidence in its reliability and accuracy. The ability of this modelling approach to explain primary and secondary metabolism makes it a valuable tool for designing novel, more efficient, and effective fermentation processes, which could have far-reaching implications in industrial biotechnology.