A Scale-Down Study of the Impact of Different Stress Parameters On the Growth and Productivity of GSCHO Cell Culture

Wednesday, November 10, 2010: 1:10 PM
Grand Ballroom E (Marriott Downtown)
William H. Scott1, Colin R. Thomas1, Chris J. Hewitt2, Gareth Lewis3 and Alvin W. Nienow1, (1)Chemical Engineering, University of Birmingham, Birmingham, United Kingdom, (2)Chemical Engineering Department, Loughborough University, , Loughborough, United Kingdom, (3)Cell Science, MedImmune, Cambridge, United Kingdom

This work aimed to study the effect of potential large-scale bioreactor environmental heterogeneities during fed-batch culture on the performance of an industrially-relevant GSCHO cell culture, producing an IgG antibody. Heterogeneity was created by applying, for the first time in animal cell culture, a two-compartment scale-down model, using a combination of a well-mixed stirred tank reactor (STR) and plug flow reactor (PFR). Feeding of glucose substrate and alkali for pH control to the PFR was analogous to feeding them to the liquid surface of a large-scale bioreactor. The flow rate through the PFR was controlled with a peristaltic pump so that the mean residence time in the PFR was equal to typical mixing excursions at the ~ 20 m3 scale. In this way, perturbations created in the PFR were analogous to perturbations created by poor mixing typical of the large-scale. The results obtained were compared with controls in which either just the STR was used; or in which circulation took place but all feeds were introduced directly into the STR. In addition to the standard parameters, antibody titre and quality were measured and flow cytometry was used to indicate cell viability and the mode of cell death. For the latter, viability was monitored by dual staining with Calcein-AM and Propidium Iodide (Calcein-AM/PI) and the mode of death by dual staining with Annexin V conjugated to phycoerythrin and Sytox Green (AV-PE/SG). Typically growth was continued for ~ 20 days and for the STR/PFR runs, this required continuous pumping for this time with a peristaltic pump. Durable neoprene was used in the pump head and the PFR tube was silicone. The results essentially fell into two categories; those without circulation and those with it. In all cases with recirculation whether nutrients and alkali were added into the STR or the PFR, significantly reduced culture duration (by ~ 48 hours) and antibody titre (~ 20% reduction) were found compared to those runs without circulation. All other key process indicators were the same for all cases, with or without recirculation, including death by necrosis. The equivalence of antibody quality in even those cases with greatly reduced viability provided strong evidence for robust antibody production in this cell line. Clearly in this study, it was not possible to conclude anything concerning the impact of bioreactor heterogeneities with this cell line. On the other hand, damage associated with peristaltic pumping has relevance to the many aspects of cell culture processes that require transfer of cells in suspension; for example, inoculation and harvest steps. For large culture volumes, pumping duration may be of significant duration. It is considered that the ‘squeezing' motion by which peristaltic pumps cause flow may impose sufficient mechanical stress on the cells to cause the relatively poor performance. It is of course possible that it may be due to long term chemical leaching or other features of the flow loop but with the high biocompatibility of the materials used in this study, this reason is considered to be less likely.

Extended Abstract: File Not Uploaded