Comprehensive characterization of an inducible mammalian expression system using 13C-metabolic flux analysis

Mammalian cell cultures have become the predominant production platform in today's multibillion dollar biopharmaceutical industry. Mammalian cells have the unique capacity to synthesize recombinant proteins with proper folding and post-translational modifications. Due to the increasing demand for biotherapeutics, the pressure to reduce the costs associated with process development and the need to shorten the time to market, continuous efforts have been made in mammalian cell technologies to maximize both the product yields and quality. This has been mostly achieved through the design of enhanced expression vectors, improved medium formulations and the application of fed-batch strategies supporting high cell densities. In recent years, there has been also a growing interest in the development of inducible mammalian expression systems, which allow decoupling of the growth and production phases and open up the possibility to define specific operating conditions that are favorable for each phase. Moreover, cells can presumably be grown more rapidly before induction, as they do not have the extra metabolic burden associated with recombinant protein expression. In such biphasic systems, a suitable timing for induction and an efficient feeding protocol pre/post-induction are among the most crucial factors to ensure a high volumetric productivity. Due to the complex metabolism of mammalian cells, the targeted optimization of cell culture processes remains challenging, but can be greatly facilitated with the aid of new metabolic engineering approaches. In particular, the use of isotopic tracers and 13C-metabolic flux analysis ( 13C-MFA) was shown to provide a more detailed and accurate description of cellular physiology than the classical metabolite balancing technique, although its application to mammalian cells is still in its infancy.;First, this method was used to analyze the central carbon metabolism of a recombinant CHO cell line in relation with cellular productivity. The goal was to quantitatively evaluate the burden imposed by recombinant protein expression on the primary metabolism of cells in culture. This was achieved by conducting a comparative study of the intracellular flux map distribution with and without the induction of recombinant protein expression. We performed cultures with multiple 13C-labeled substrates and measured the resulting mass distribution of extracellular metabolites (lactate and three amino acids) by mass spectrometry. This additional data was used in conjunction with extracellular rate measurements to obtain reliable metabolic flux estimates. Through this analysis, we have notably demonstrated that heterologous protein expression is correlated with small, but significant differences in a number of important intracellular pathways, most of them related to ATP and NADPH formation, including the pentose phosphate pathway, the malic enzyme reaction and the TCA cycle. Induced cells were found to exhibit an increased flux of pyruvate into the TCA cycle, indicative of a more efficient glucose utilization. However, no significant difference could be inferred with respect to amino acid metabolism, including glutamine. Since a temperature shift was performed at the time of induction, this study was also the first comprehensive characterization of CHO cell metabolism under mild-hypothermia conditions.;We have then investigated the impact of the timing of induction on culture productivity, since this parameter can affect both the cumulative biomass concentration and the cell specific productivity. Cells taken at different stages of growth were transferred and induced in fresh medium and the kinetics of growth, nutrient consumption and product formation were compared during the production phase. Compared to cultures induced at a high cell concentration, low cell density inductions achieved lower maximum cell concentrations, but exhibited higher cell specific productivity and greater culture longevity, which gave a greater final product titers. To gain more physiological insights into the observed differences, 13C-metabolic flux analysis was performed to characterize and compare the metabolism of cells induced at respectively low and high cell densities. The range of measurements was extended by including the mass distributions of two organic acids. In line with differences observed in terms of growth and productivity, several calculated intracellular fluxes were found to be affected by the timing of induction. It was inferred that the corresponding availability of nutrients had a determinant impact on the induction phase. Most notably, glucose utilization efficiency was increased in high cell density induction, whereas the contribution of amino acids became marginal.;The previous results were however obtained by doing a complete medium exchange at the time of induction, an operation which cannot be easily implemented at large-scale. For this reason and since the results implied a strong correlation between nutrient availability and cellular growth/productivity, we then explored the possibility to increase the product yield by using a fedbatch process. We found that performing concentrated feed additions post-induction greatly increase the productivity of cultures compared to induction in batch mode. More interestingly, we observed that when glutamine was fed in excess to the cells, the cultures reached greater maximum cell concentrations. While cell growth was comparatively lower in glutamine-limited fed-batch, the final product titers were found to be similar, indicative of an increased cell specific productivity. To further assess the physiological impact of glutamine levels on the cells, a comparative 13C-metabolic flux analysis was conducted in semi-continuous cultures to analyze both culture regimes. The marked changes in cellular growth and productivity were clearly reflected in the metabolism of the cells. In particular, lactate metabolism was found to be greatly influenced by the amount of glutamine present in the feed during the production phase.;The main objective of this work was thus to apply 13C-metabolic flux analysis to quantitatively characterize the primary metabolism of recombinant CHO cells harboring an efficient inducible expression system, as a systematic approach to guide the determination of key process conditions and allow the rational development of a feeding protocol to increase culture productivity. To this end, analytical methods for conducting informative labeling experiments and mass spectrometry measurements, as well as computational tools allowing reliable metabolic flux quantification had to be developed or adapted from the literature.;Overall, the results from all our studies tend to suggest that there may be some negative correlation between cell growth and cellular productivity. Nonetheless, the metabolic characterizations performed in this work may provide useful clues for the identification of physiological markers of cell growth and/or productivity that could further guide the optimization of inducible expression systems. Our work also demonstrates that 13C-metabolic flux analysis is a valuable tool for characterizing cell metabolism and supporting cell culture process optimization.