Lawrence Berkeley National Laboratory

Scaling biomanufacturing from laboratory to industrial scale poses significant challenges, especially for continuous fermentation. This study investigates these challenges using a β-carotene-producing Yarrowia lipolytica strain. Through fermentation experiments and proteomics, we have assessed how fermentation modes, carbon sources, dissolved O₂, and media composition influence long-term bioproduction. In shaking flask subcultures, the strain maintained β-carotene production for over ~30 generations. However, in continuous fermentations, subpopulation shifted toward faster-growing low-producers, leading to significant production losses within just ~18 growth generations. This process was accelerated by O₂ limitation and high bioreactor dilution rates. Using canola oil as a carbon source increases population heterogeneity but enhances β-carotene biosynthesis and prolongs production compared with glucose-based media. Kinetic modeling suggests that strains optimized for the highest production in laboratory settings may be less robust in industrial environments, where suboptimal yet faster-growing variants gain a competitive edge under prolonged stress and ultimately shape overall continuous fermentation performance.