(94e) CHO Cell Culture Process Development Via Metabolic Modeling

Biopharmaceutical processes development can involve expensive and laborious experiments involving screening of cell clones, formulation of cell culture media, optimization of bioreactor parameters, and developing feed media supplementation strategies [1]. Leveraging simulations of bioprocess outcomes via process models can reduce the number of experiments required to develop bioprocesses [2]. However, the failure to capture the complexity of cellular metabolism in these models can lead to unreliable predictions. To overcome these limitations, a dynamic metabolic flux analysis (DMFA) model was developed by integrating a reaction network (70 reactions describing amino acid metabolism, glycolysis, and TCA cycle) with a Monod based kinetic model to simulate the dynamic concentration profiles of glucose, lactate, amino acids, monoclonal antibody (mAb) titers, and ammonia when provided with culture parameters such as bioreactor seeding density, initial media concentrations, feed media supplementation details, and bioreactor pH. The model was trained by using data from fed-batch experiments [3].

Model-based process development was demonstrated via three experimentally validated case studies. First, the DMFA model was used to predict intensified fed-batch bioreactor performance by predicting the impacts of high bioreactor seeding density and bioreactor pH on the viably cell density and final mAb titers. The predictions of amino acid metabolism provided insights on depletion or accumulation of various amino acids. Thus, providing mechanistic information that is needed to rebalance media composition. Second, the metabolic model that was trained on only fed-batch data was used to predict the impact of perfusion bioreactor process parameters such as perfusion rate and bleed rate on viable cell density, mAb titers, and metabolite concentrations. Thus, demonstrating that existing fed-batch data can be leveraged to predict perfusion bioreactor performance. Finally, we evaluated the impact of AMBIC basal and feed media on cell culture performance. The model not only successfully predicted the cell culture performance but also provided valuable information on depletion of certain amino acids. This information can lead to improved media composition and feed media supplementation schedules.

References

1. Bielser, J.-M., et al., Perfusion mammalian cell culture for recombinant protein manufacturing – A critical review. Biotechnology Advances, 2018. 36(4): p. 1328-1340.
2. Sommeregger, W., et al., Quality by control: Towards model predictive control of mammalian cell culture bioprocesses. Biotechnol J, 2017. 12(7).
3. Reddy, J.V., et al., Flux balance analysis and peptide mapping elucidate the impact of bioreactor pH on Chinese hamster ovary (CHO) cell metabolism and N-linked glycosylation in the fab and Fc regions of the produced IgG. Metabolic Engineering, 2025. 87: p. 37-48.