On-line monitoring of physiological parameters of insect cell cultures during the growth and infection process

On-line monitoring df insect cell cultures used for the production of recom binant proteins with the baculovirus expression vector system (BEVS) provid es valuable tools for the optimization, operation, and control of the produ ction process. The relative permittivity (epsilon') and CO2 evolution rates (CER) were measured on-line using the biomass monitor and the infrared CO2 analyzer, respectively. The growth and infection phases of two different c ell lines, Spodoptera frugiperda (Sf-9) and Trichoplusia ni (High-5), were monitored using the above measurements. These in turn were correlated to th e progress of the culture by using the off-line measurements of protein pro duced, virus titer, and biovolume, which is the product of viable cell dens ity and mean cell volume. The epsilon', CER, and the biovolume profiles wer e closely matched during the growth phase of cells when grown in a batch or fed batch culture. The relationship became more complex when the cultures were either in stationary phase;or in the postinfection phase. The epsilon' profile was found to be a good indicator of the process of synchronous bac uloviral infection, showing a plateau between 18 and 24 h postinfection (hp i), the period during which budded virus is produced, and a peak at approxi mately 48 hpi correlated to the onset of accelerated cell lysis. The CER pr ofile continues to increase after the growth period with a peak around the 24 hpi period, after which there is a decline in the profile corresponding to release of virus as seen from virus titer determinations. This was exami ned for Sf-9 cultures under conditions of cell densities from 3 to 50 x 10( 6) cells/mL and MOI values ranging from 0.001 to 1000. The profiles were fo und to be similar also in the case of the High-5 cells. Thus both measureme nts give reliable information regarding the physiological status of the cel ls as seen from their correlation to virus and protein production. A furthe r combination of these with the off-line measured parameters such as the bi ovolume and metabolite concentrations can give a more detailed understandin g of the process and help in the better design and automation of these proc esses.