(482e) Analyzing Protein Glycosylation in Mammalian Perfusion Cell Culture Using Glycosylation Flux Analysis

Hutter, S., ETH Zurich
Karst, D., ETH Zurich
Schweigler, T., ETH Zürich
Morbidelli, M., ETH Zürich
Gunawan, R., ETH Zurich
In recent times, perfusion cultures of mammalian cells, particularly Chinese hamster ovarian (CHO) cells, have been drawing a growing interest as an avenue for continuous biopharmaceutical production. Perfusion cultures provide several benefits over traditional fed-batch processing such as higher yield and improved product quality. In the production of monoclonal antibodies (mAb), N-linked glycosylation is among the most important product quality attributes. Such post-translational modification has strong effects on the biological activity and therapeutic efficacy of these drugs. While the macroscopic process variables such as the viable cell density and the concentrations of main metabolites and products in the media are measured throughout the cell culture operation, there is little or no direct measurements on intracellular variables. For this reason, model-based data analysis such as metabolic flux analysis is often required to understand intracellular changes during cell culture production.

In this work, we extended the traditional metabolic flux analysis (MFA) to study N-linked glycosylation process during the operation of a perfusion CHO cell culture reactor to produce immunoglobulin gamma (IgG) antibodies. Like MFA, this novel analysis, that we called glycosylation flux analysis (GFA), is a flux-based technique that relies on a stoichiometric model of the reaction network in the protein glycosylation to predict reaction rates or fluxes, under the assumption of pseudo steady-state condition. In dynamic GFA (DGFA), the steady-state assumption is relaxed by allowing fluxes to vary linearly between time points, as done in dynamic MFA [1]. We employed GFA as well as DGFA to link the extracellular measurements of cell viability, IgG titer and fractions of glycoforms to the intracellular changes in the glycosylation reaction network during long term continuous cultivation.

The overall goal of our study was to understand the regulation of the glycosylation process in CHO cells during perfusion culture, particularly during the initial phase and when going through transitions over different operating conditions. The results of (D)GFA showed that while macroscopic variables such as the viable cell density could be controlled at a desired set-point, intracellular processes such as glycosylation varied over time. This time dependent change in fluxes through the glycosylation network suggested that the viable cell density might not be the most appropriate controlled variable for optimizing the perfusion culture performance while maintaining a constant product quality. By computing flux conversions, defined as the ratio between an efflux and the total influx around a glycoprotein in the network, we linked changes in the activities of the pertinent enzymes to external macroscopic variables [2]. The acquired knowledge about the dynamics of glycosylation process and the enzymes involved will be applied for further process optimization and the control of mAb product quality.

[1] Leighty, R.W., Antoniewicz, M.R., 2011. Dynamic metabolic flux analysis (DMFA): a framework for determining fluxes at metabolic non-steady state. Metab. Eng.

[2] Hang, I., Lin, C., Grant, O.C., Fleurkens, S., Villiger, T.K., Soos, M., Morbidelli, M., Woods, R.J., Gauss, R., Aebi, M. (2015.) Analysis of site-specific N-glycan remodelling in the ER and the Golgi. Glycobiology.