A high-throughput assay to assess enzyme activity in central metabolism of production cell lines

Mammalian cells are widely used for the production of a variety of biopharmaceuticals, such as monoclonal antibodies, recombinant proteins, and viral vaccines. However, during growth in glutamine-containing media, large amounts of toxic by-products, such as lactate and ammonia, are secreted into the culture medium that often not only affects cell viability, productivity and product quality but also can prevent growth to high cell densities. Many different analytical methods, such as liquid chromatography-mass spectrometry for measuring intracellular metabolites, were used to characterize and to better understand the inefficient metabolism of production cells. The objective of the present study to further assess cell metabolism was the development of an efficient assay system based on 96-well microplates for central metabolic enzyme activities, which allowed the grouping of various enzymes in modules that share a common detection method. The enzyme platform consists of four sensitive cycling assays to quantify low amounts of NAD+(H), NADP+(H), glycerol 3-phosphate or dihydroxyacetone phosphate, and glutamate or 2-oxoglutarate. The sensitivity limit of all cycling assays was between 0.025 and 0.4 nmol product. Cell extracts could, therefore, be highly diluted, which reduced eventual interferences caused by other components in the extract and additionally minimized under- or overestimates of actual enzyme activity. Furthermore, possible enzyme inhibition by high concentration of a product was prevented since substrate concentrations could be maintained at a near constant level throughout the assay. Adherent Madin-Darby canine kidney (MDCK) cells were grown to stationary and exponential phases in 6-well plates in GMEM supplemented with glutamine or pyruvate, and key metabolic enzyme activities of cell extracts were analyzed. Significant differences were found in maximum enzyme activities from cells grown with pyruvate-containing medium compared to glutamine-containing medium. The overall activity of the pentose phosphate pathway was up-regulated during exponential cell growth in pyruvate-containing medium, which suggests that more glucose 6-phosphate was channeled into the oxidative branch. Furthermore, the anaplerotic enzymes pyruvate carboxylase and pyruvate dehydrogenase showed higher cell specific activities with pyruvate. Increased specific activities were also found for NAD+-dependent isocitrate dehydrogenase, glutamate dehydrogenase and glutamine synthetase in MDCK cells grown with pyruvate. It can be assumed that the increase in enzyme activities was most likely required to compensate for the energy demand and to replenish the glutamine pool. On the other hand, the activities of the glutaminolytic enzymes aspartate transaminase, alanine transaminase, malic enzyme (ME) and phosphoenolpyruvate carboxykinase were decreased in cells grown with pyruvate, which seems to be related to a decreased glutamine metabolism. When MDCK cells were infected with an influenza A (H1N1) virus at a high multiplicity of infection, infected cells showed an up-regulation of some key enzymes producing the reduction equivalent NADPH (glucose-6-phosphate dehydrogenase, 6-phosphogluconate dehydrogenase, and ME) and acetyl-CoA (citrate lyase and acetate-CoA ligase), a precursor needed for lipid and cholesterol biosynthesis. It seems that the synthesis of fatty acids and cholesterol plays a crucial role for the replication of influenza viruses in adherent MDCK cells as they acquire lipid envelopes from their host cells. Based on the established enzyme assays the metabolic states of production cell lines can now be further characterized. This can then be used to improve our understanding of metabolic pathways relevant for cell line and media optimization. Furthermore, it will support the validation of mathematical models of cellular metabolism in systems biology approaches