Conference Dates

November 1-5, 2015


High cell density perfusion has the potential to generate intensified processes. A target for such perfusion process can be to maintain a steady state culture around 100 x 106 cells/mL. A challenge for the development of these processes is to minimize the medium consumption, to reduce the costs associated with this raw material and with the harvest processing. This is particularly true for processes at densities such as 100 x 106 cells/mL for which several reactor volumes per day are typically necessary.

We have shown that cell specific perfusion rate (CSPR) is an approach, which has proven to enable densities of CHO cells above 200 x 106 cells/mL. To appreciate the impact of the cell density on the metabolic landscape in an antibody producing CHO cell perfusion culture using CSPR, multivariate data analysis of extracellular metabolomic data generated by high-resolution liquid chromatography–mass spectrometry (LC–MS) was carried on. We observed that the extracellular metabolic profile was by far more stable in perfusion mode than in fed-batch mode, despite a significantly larger range of cell densities in the former case. There was a strong correlation between the composition of the exometabolome and the viable cell density between 8 and 207 x 106 cells/mL. The cell specific glutamine consumption rate as well as, to some extent, the cell specific glucose consumption rate were only correlated with the composition of the exometabolome when the viable cell density was above 127 x 106 cells/mL. No correlations between the exometabolome and the cell specific productions of lactate, ammonia or antibody were observed. These results show that the metabolism of the cells was very stable at densities lower than 127 x 106 cells/mL. Above this density, the metabolic footprint analysis revealed variations of the cellular metabolism: in particular modifications associated with the glutathione metabolism were observed.

We addressed the CSPR minimization by sequentially varying the cell density and/or the perfusion rate in a culture stabilized around 30 x 106 cells/mL by cell bleeds. In our system, we observed that the cell specific antibody production rate was independent of the CSPR below 52 pL/cell/day and positively correlated with this parameter above this value. Furthermore, the cell specific glucose consumption and lactate production rates increased with the CSPR. Interestingly, during the CSPR optimization, we showed that increasing the cell density was not systematically equivalent to reducing the perfusion rate despite the fact that the CSPR is equal to the ratio of the perfusion rate by the cell density. This was due to a substrate threshold determining the occurrence of different glucose consumption kinetics.