Carbon dioxide control in bioreactors and the application of principal component analysis to cell culture process data

Abstract

Carbon dioxide levels can vary with bioreactor scale in mammalian cell cultures but its effects on cell culture process is poorly understood. Varied results have been presented in the literature on the effects of pCO2 on the process performance of mammalian cell cultures. If pCO2 control can be demonstrated to show that it has the potential to effect or improve cell culture performance then the control at production scale and during scale-up would be beneficial. The primary objective of the research was to examine the effect of carbon dioxide concentration (pCO2) on a monoclonal antibody (mAb) producing Chinese Hamster Ovary (CHO) cell line in order to assess if pCO2 is a critical process parameter. During the research investigations, increasing pCO2 control resulted in an associated increase in osmolality; therefore, the role of the osmolality increase was also investigated. The main process responses examined were the productivity and product quality of a cell culture process, as well as the effect of pCO2 on the cell metabolism. The effect of pCO2 control from low to high (20 to 100 mmHg) and at elevated levels (140 mmHg) was examined and compared to a baseline condition (no pCO2 control). Results indicated that both increasing pCO2 and osmolality level independently led to increased product titre. Operation at low to high pCO2 gave reduced product quality attributes, compared to the baseline. Elevated pCO2 control resulted in equivalent product quality attributes compared to the baseline. Carbon dioxide level had the greatest impact on the product quality attributes, with osmolality playing little or no role. In summary, pCO2 control level was shown to be a critical process parameter and should be considered in quality by design, small scale optimisation and scale-up studies. The current research results disagreed with a selection of the varied results reported in the literature, on the effect on pCO2/osmolality on a mammalian cell culture process. In previous studies non-optimised process controls and conditions were used, such as the use of unrepresentative reactor vessels (e.g. T-flasks). This would lead to interacting effects from other process parameters. The majority of studies of studies on the effects of pCO2/osmolality carried out were conducted over 10 years ago. Since then, advances in process monitoring and control, media formulation, cell lines, analytical techniques and large increases in product titres, have occurred. Principal component analysis (PCA) was applied, with the aim of extracting further understanding from the data set generated in the pCO2 study. PCA models demonstrated grouping of pCO2 control and osmolality levels due to differences in process variables, relating to the control of pCO2. Overall, the PCA results statistically confirmed that pCO2 control leads to operational differences in the cell culture performance.