Process intensification for recombinant protein production in E. coli via identification of active nodes in cellular metabolism and dynamic flux balance analysis

Complex media supplemented with a carbon source are commonly used in bioprocesses for recombinant protein production in E. coli. Optimizing these processes is challenging and requires precise understanding of cellular metabolism and nutrient requirements. Compared to a Design of Experiments approach that necessitates extensive experimentation, metabolic modeling using a genome scale metabolic model (GEM) offers a more predictive and systematic approach to guide process optimization by identifying specific metabolic bottlenecks. In addition, spent media analysis (SMA) can unravel the preferential utilization of different media components during the bioprocess. Here, we integrated the updated E. coli GEM with time course SMA data from a fed-batch process and performed dynamic flux balance analysis (dFBA) to identify metabolites that function as active nodes and are vital for cellular function. These are potential target supplements to boost cellular activity and in turn the recombinant protein productivity. Using an iterative approach of performing fermentation, SMA, and metabolic modeling, we optimized the bioprocess in just five experimental trials, resulting in a six-fold increase in protein productivity. The optimized feeding strategy involved yeast extract with amino acid supplementation (Ser, Thr, Asp, Glu) and increased oxygen transfer rates. This approach demonstrates significant promise for application in bioprocess optimization.