PUB - Publikationen an der Universität Bielefeld

Hyperosmolality can occur within the volume of non-ideally mixed large-scale bioreactors during biopharmaceutical manufacturing. It negatively impacts cellular growth and can inhibit productivity, which significantly decreases the robustness and reproducibility of the processes through their up- and downscaling. Surprisingly, hyperosmolality often leads to an increased product formation rate and positively influences product quality. To study the mechanisms through which the hyperosmolality influences the CHO cells, we conceptualized a set-up where an industrial over-supplemented feed was used as an active osmotic agent, controlled for its specific effects by including a mannitol-supplemented feed to reach the same osmolalities and compare these two treatment conditions. We investigated the whole range of osmolalities tolerated by the CHO cells without an immediate viability drop (+70 to +230 mOsm/kg). We detected a highly significant increase in mitochondrial membrane potential, mitochondrial mass, and mtDNA copy number per cell in oversupplemented feed and mannitol conditions (compared to control cultivated under physiologic osmolality). However, when cell volume was considered, all three mitochondrial parameters increased to a much lower extent than the cell’s volume. The latter is almost triplicated in our experiments. Therefore, mitochondria must be diluted in the volume of large cells exposed to hyperosmotic stress. We also detected a significant decrease in mitochondrial membrane potential, a key parameter correlating with mitochondrial fitness and ATP production capacity, if the exposure to harsh mannitol hyperosmolality (530 mOsm/kg) and oversupplemented feed (460 and 530 mOsm/kg) were prolonged in their duration (4 days). By conducting a label-free quantitative proteome analysis to compare the protein abundances between the feed-exposed vs. control cells, we detected 186 molecular players involved in osmotic stress response. Among these, a massive activation of unfolded protein response (UPR) mediators, heat-shock proteins involved in DNA repair, overexpression of proteins involved in the glutathione detoxification system, and peroxidases confirmed that feed exposure leads to a high degree of oxidative stress, the underlying cause of the hyperpolarization of the mitochondrial membrane and an inducer of random DNA damage. Second, cells overexpress several members of the septin protein family. Septins increase the mechanical stability of the cell by interacting with the cellular membrane and probably forming compartments within the cell, which may segregate misfolded proteins from those which still retain their activity and function. This might be a hitherto unknown adaptation strategy to hyperosmolality. Third, we detected substantial rearrangements in the extracellular matrix (ECM) and secretome of CHO cells. Several otherwise abundant ECM components, such as collagens, laminins, and nidogens, are significantly downregulated in feed cells. However, although some proteins are markedly altered in their abundance, the overall protein composition of the cell seems to be retained to a large extent, i.e., the feed-exposed cells are made of the same “stuff”, i.e, contain the same basic protein inventory, as control cells, except for the key alterations mentioned above. At last, we detected and characterized a significant heterogeneity of cellular response if confronted by a hyperosmotic stimulus. Upon exposure of CHO cells to both nutrient-over-supplemented and mannitol-supplemented feeds, some cells seem to be able to retain an almost average proliferation speed. On the contrary, some ablate proliferation and only increase in size. Here, the already prerequisite clonal metabolic variations, known for CHO cells, and the cycle phase in which each cell was in upon osmolality change seems to be decisive. The knowledge acquired for suspension-grown CHO cells and presented in this work can be used to identify the potential of hyperosmotic exposure to affect the cellular performance of producing CHO lines concerning both the quantity and quality of the product. This could be realized and evolved by informed feeding strategy design and optimization of biopharmaceutical production processes.