(173g) Analyzing Metabolic Shifts in Hybridoma Cells Using Dynamic Metabolic Flux Analysis Considering Network Uncertainty

Mammalian cells, such as CHO and hybridoma cells, play an important role in the production of monoclonal antibody (mAb) drugs. The need for increasing the productivity of mammalian cell culture while maintaining the quality of mAb product demands a deeper understanding of the cell metabolism. To this end, mathematical models of metabolic networks have been used for gaining insights and predicting metabolic flux distribution. Two constraint-based methods, namely flux balance analysis (FBA) and metabolic flux analysis (MFA), have been developed for the above purposes based on pseudo-steady state assumption (PSSA) [1]. Extension of these methods to dynamical analysis, i.e. dynamic FBA (dFBA) and dynamic MFA (dMFA), are also available [2, 3].

Both dFBA and dMFA use PSSA. The validity of such an assumption and the outcome of the analysis depend on having an accurate description of the metabolic network. One of the advantages of FBA is the ability to analyze large network models. Therefore the issue of network accuracy above can be mitigated by considering genome-scale metabolic networks. However, the outcome of FBA depends on the assumed cellular objective function. On the other hand, by considering only a subset of metabolic reactions that are important, MFA does not require any assumption on cellular objective function. However, the validity of PSSA for the metabolic subnetwork under consideration is uncertain.

In this work, we develop an extension of dMFA [3] to address uncertainty associated with using PSSA. The method is based on relaxing the PSSA and provides a guaranteed upper bound on errors in the steady-state condition Sv = 0, where S is the stoichiometric matrix and v is the metabolic fluxes. More specifically, we allow small flux imbalance in the metabolic network model as long as there is a significant improvement in the data fitting (using Akaike Information Criterion). We have applied the dMFA with uncertainty (dMFAu) to analyze data from hybridoma cell culture production of IgG (CRL 1606), with the goal of predicting metabolic shifts under different pH conditions. Based on the analysis, we explain changes in the hybridoma metabolism during the transition from lactate production to lactate consumption.

[1] Antoniewicz MR (2013) Dynamic metabolic flux analysis--tools for probing transient states of metabolic networks. Curr Opin Biotechnol, 24(6): 976-978.

[2] Mahadevan R, Edwards JS, Doyle FJ 3^rd(2002): Dynamic flux balance analysis of diauxic growth in Escherichia coli. Biophys J 83(3):1331–1340

[3] Leighty RW, Antoniewicz MR (2011) Dynamic metabolic flux analysis (DMFA): a framework for determining fluxes at metabolic non-steady sate. Metab Eng 13: 745–755.