Driving Consumption of Amino Acids
Although amino acid supplementation for the high nutrient condition had minimal impact, the larger increases in cell growth and productivity within Criterion 2 demonstrated an approach to identify limiting nutrients in the moderate and low nutrient controls. However, to determine if the limiting nutrients were indeed the model derived amino acids, consumption rates were compared between model conditions and the respective control within each nutrient feed level from Criterion 2 conditions (Fig. 8). The time dependent differences in amino acid consumption rates between the process controls for each nutrient feed level were represented as either a positive (green) difference reflecting an increase in consumption or a negative (red) difference reflecting a decrease in consumption (Fig. 7). For classification purposes, the amino acids supplemented to the model-driven conditions were highlighted and were grouped as either glucogenic or ketogenic. Glucogenic amino acids can provide a carbon source to the cells to biosynthesize glucose for metabolism whereas ketogenic amino acids provide a carbon source to generate ketone bodies which can serve as a secondary energy source in oxidative metabolism (Brosnan, 2003; Noguchi et al., 2010).
Previously, it has been postulated that amino acid consumption is largely dependent on the concentration of the corresponding amino acid (Fan et al., 2015). Indeed, this was true for the high nutrient feed conditions in which not only the model-supplemented amino acids, but almost all amino acids showed an increase in consumption up to and including day 11 for both models. The few with reduced consumption included glutamic acid, which showed a marked decrease in consumption across all time points after day 3. Although glutamic acid is a nonessential amino acid for CHO cells, we posit that this decrease in glutamic acid consumption is driven by a shift to increase consumption of other supplemented nonessential amino acids, such as glycine and alanine. However, due to the limited increase in cell growth and mAb productivity from these conditions, overall increased consumption did not shift the cells to a higher metabolic state, suggesting a saturation state for cells and a point of maximum contribution by amino acids.
In contrast, moderate and low nutrient conditions showed a more drastic difference in the consumption patterns when compared to the high nutrient conditions. Most noticeably, there was a marked decrease in the consumption of ketogenic amino acids and increased consumption of glucogenic amino acids suggesting a greater need for energy and biosynthesis due to the limited nutrient conditions. Furthermore, for moderate nutrient conditions the growth model-derived amino acids had increased consumption earlier in the culture and production model-derived amino acids such as methionine were consumed more in the latter half of the culture, supporting the distinct difference in metabolic requirements between the cell growth phase and the high mAb-producing stationary phase. In the low nutrient condition, there was a decrease in majority of the amino acids within the growth model but increase in glutamic acid and alanine in the latter half of the culture. In the low nutrient production model, methionine and eventually lysine showed increased consumption resulting in increased mAb productivity. In all cases, model-derived amino acids were generally consumed more when supplemented according to Criterion 2 than process control cultures for each respective nutrient condition. Accordingly, these amino acids were potential limiting nutrients that pushed cells into a higher metabolic state when supplemented.