High-throughput fermentation screening for the yeast Yarrowia lipolytica with real-time monitoring of biomass and lipid production

Alexandre Back, Tristan Rossignol, François Krier, Jean-Marc Nicaud, Pascal Dhulster

Because the model yeast Yarrowia lipolytica can synthesize and store lipids in quantities up to 20 % of its dry weight, it is a promising microorganism for oil production at an industrial scale. Typically, optimization of the lipid production process is performed in the laboratory and later scaled up for industrial production. However, the scale-up process can be complicated by genetic modifications that are optimized for one set of growing conditions can confer a less-than-optimal phenotype in a different environment. To address this issue, small cultivation systems have been developed that mimic the conditions in benchtop bioreactors. In this work, we used one such microbioreactor system, the BioLector, to develop high-throughput fermentation procedures that optimize growth and lipid accumulation in Y. lipolytica. Using this system, we were able to monitor lipid and biomass production in real time throughout the culture duration. The BioLector can monitor the growth of Y. lipolytica in real time by evaluating scattered light; this produced accurate measurements until cultures reached an equivalent of OD600nm = 115 and a cell dry weight of 100 g L(-1). In addition, a lipid-specific fluorescent probe was applied which reliably monitored lipid production up to a concentration of 12 g L(-1). Through screening various growing conditions, we determined that a carbon/nitrogen ratio of 35 was the most efficient for lipid production. Further screening showed that ammonium chloride and glycerol were the most valuable nitrogen and carbon sources, respectively, for growth and lipid production. Moreover, a carbon concentration above 1 M appeared to impair growth and lipid accumulation. Finally, we used these optimized conditions to screen engineered strains of Y. lipolytica with high lipid-accumulation capability. The growth and lipid content of the strains cultivated in the BioLector were compared to those grown in benchtop bioreactors. To our knowledge, this is the first time that the BioLector has been used to track lipid production in real time and to monitor the growth of Y. lipolytica. The present study also showed the efficacy of the BioLector in screening growing conditions and engineered strains prior to scale-up. The method described here could be applied to other oleaginous microorganisms.