Understanding and overcoming physiological stress responses in Yarrowia lipolytica to dissolved O2 and CO2 in fermentation processes

Society is facing several major challenges due to a significant reliance on fossil-based resources. Bioprocess engineering, an evolving field with advanced technologies, offers a wide range of potential solutions through the use of renewable feedstocks and microbial cell factories. As such, a cleaner production of everyday needs like chemicals, energy, and food can be achieved. A special microbial host, and focus of this Ph.D. thesis, is the yeast Yarrowia lipolytica. Due to its potential to grow on a broad range of renewable carbon sources and ability to produce a variety of relevant compounds, interest in the microbe has been rising over the years. This has led to an increased availability of metabolic engineering tools and the development of several industrial processes.

Although there are plenty of great initiatives, many concepts struggle to make it to an industrially sized and commercially viable level. As many of these bioprocesses are developed in laboratories with a strong focus on optimizing pathways, the topic of industrial robustness is often overlooked. Then, as the process transitions into a large scale, unexpected losses in yield and productivity are observed. Successful scale-up of bioprocesses from lab (< 300 L) to production scale (>20 m3), requires a precise understanding of the microbial response to the harsh and heterogeneous environment that the microbes experience on an industrial scale.

Therefore, to make bioprocess engineering more efficient, a couple of topics need more detailed assessment. First, an improved understanding of the industrial vessels is required. In this Ph.D. thesis, this is done by modelling reactor volumes up to 90 m3 to understand the extent of heterogeneities cells experience regarding substrate, oxygen, and carbon dioxide. These simulations set a framework for so-called scale-down experiments, in which Y. lipolytica is exposed to relevant conditions in the laboratory. The metabolic/regulatory response is then studied by multi-omics data integration.
Further, in this thesis, to advance our understanding of Yarrowia’s regulatory response, a high quality RNA-seq database has been generated under different conditions. The results are used in a machine learning approach to further unravel and structure the regulatory network of Y. lipolytica. This allows for the identification of groups of genes that are co-regulated in response to a stimulus. This so-called iModulon approach has been successfully applied to different prokaryotes, but this is the first of its kind for a eukaryotic strain.

The results presented in this work set a framework that can be followed for any type of organism to achieve a more streamlined bioprocess development. More specifically, this Ph.D. thesis provides a broad knowledge base of Y. lipolytica in response to industrial bioreactor heterogeneities. Finally, this study allows us to focus on further strain engineering approaches towards making Yarrowia a robust fermenterphile for industrial applications.