Evaluation of predictive tools for cell culture clarification performance.

Abstract

Recent advances in the productivity of industrial mammalian cell culture processes have resulted in part in increased cell density. This increase and the associated increase in cellular debris are known to challenge harvest operations, however this understanding is limited and largely qualitative. Part of the issue arises from the heterogeneous size and composition of cellular debris, which makes harvest feed stream extremely difficult to characterize. Improved characterization methods would facilitate the development of clarification approaches that are consistent and scalable. This work describes how both particle size and cholesterol analysis can be used to characterize the feed stream. Particle size analysis by focused beam reflectance and dynamic light scattering are shown to be predictive of centrate filterability under certain harvest conditions. Because of the particle size range limitations of each detector, their applicability is limited to a particular stage or method of clarification. The measurement of cholesterol present in the cell culture supernatant or centrate was successfully used in providing relative amount of lysed cellular debris and enabled us to predict clarification performance of acid precipitated harvest regardless of particle size distribution profile.