Control of gene expression in E. coli BL21 (DE3) population upon long-term continuous cultivation.

Abstract

Continuous cultivation in bioreactors offers several advantages for recombinant protein production, including higher volumetric productivity, consistent product quality, and maximized equipment uptime. However, its industrial deployment remains limited due to the emergence of population heterogeneity, especially when using microorganisms engineered with burdensome gene circuits. In this study, we investigated the population dynamics of Escherichia coli BL21 (DE3) carrying a burdensome gene circuit, the T7-based GFP expression system, during long-term chemostat cultivation (>100 hours). An automated flow cytometry setup enabled single-cell fluorescence monitoring and the assessment of population heterogeneity, using entropy as a metric, under different environmental conditions. As expected, the experiments have shown that population heterogeneity appears over time in continuous culture. To mitigate this diversification, we tested two strategies: modifying the carbon source (glucose vs. xylose) and reducing the cultivation temperature. The assumption was that lowering the strain’s maximal growth rate would reduce the switching cost and promote population homogenization, reflected by lower entropy. However, both strategies had limited impact on homogenizing GFP expression within the population, indicating that reducing switching cost alone is insufficient to control population diversification in continuous cultures. Finally, this work highlights key factors influencing population stability, including the importance of switching cost, the critical role of culture history before and during continuous culture, and the phenomenon of environmental escape. This latter mechanism allows part of the population to evade unfit phenotypic states and may also be connected to mutational escape. Controlling it could therefore help limit the emergence of mutants in continuous bioreactors. Overall, this work provides evidence that effective control of population dynamics in continuous recombinant protein production requires a multifactorial approach integrating host population traits, genetic circuit design, and environmental parameters.