(504a) Model-Based Analysis of the Monoclonal Antibody Glycosylation across CHO Cell Lines

N-linked glycosylation, a post-translational modification where the glycan is attached to the nitrogen atom of asparagine in a protein, is a critical quality attribute in biologic manufacturing because it can affect the half-life, immunogenicity, and pharmacokinetics of therapeutic monoclonal antibodies (mAbs) [1]. To maintain a consistent glycosylation profile, multiple mechanistic models have been developed to predict glycosylation profiles for both fed-batch [2] and perfusion [3] cell culture processes. While the mechanisms and associated model are expected to be consistent across cell lines, some of the specific model parameters may differ between cell lines. Despite all Chinese hamster ovary (CHO) cell lines sharing a common ancestor, extensive mutagenesis and clonal selection has resulted in substantial genetic heterogeneity among them [4].

To evaluate the influence of these genetic differences on the bioprocess and mAb quality, we establish a multiscale glycosylation model and adapt the model to various cell lines and culture conditions. Due to the complex glycan measurements and high cost of cell culture experiments, there are limited data available for model evaluation and parameter estimation. Therefore, we employ Bayesian estimation to incorporate prior knowledge from the existing literature and alleviate non-identifiability problems. The uncertainties of model parameters are then propagated to quantify the uncertainties of key model prediction variables. To improve the computational speed of parameter estimation, we employ several strategies including a decomposition method, a simplified glycosylation reaction network, and a tailor-made parallelization scheme.

With the above parameter estimation framework, we obtain and compare the probability distribution for each model parameter among various cell lines. While many parameters maintain similar distribution, some concentration and kinetic parameters of certain Golgi-resident enzymes exhibit significant changes across different cell lines. The expression of these enzymes is likely upregulated or downregulated through genetic engineering methods within cell line development. This quantitative analysis of the host cell impact on mAb glycosylation profiles provides insight into future glycoengineering strategies for improved product quality.

[1] E. Edwards, M. Livanos, A. Krueger, A. Dell, S.M. Haslam, C. Mark Smales, D.G. Bracewell, Strategies to control therapeutic antibody glycosylation during bioprocessing: Synthesis and separation, Biotechnol Bioeng 119(6) (2022) 1343-1358.

[2] P. Kotidis, P. Jedrzejewski, S.N. Sou, C. Sellick, K. Polizzi, I.J. Del Val, C. Kontoravdi, Model-based optimization of antibody galactosylation in CHO cell culture, Biotechnol Bioeng 116(7) (2019) 1612-1626.

[3] D.J. Karst, E. Scibona, E. Serra, J.M. Bielser, J. Souquet, M. Stettler, H. Broly, M. Soos, M. Morbidelli, T.K. Villiger, Modulation and modeling of monoclonal antibody N-linked glycosylation in mammalian cell perfusion reactors, Biotechnol Bioeng 114(9) (2017) 1978-1990.

[4] D. Reinhart, L. Damjanovic, C. Kaisermayer, W. Sommeregger, A. Gili, B. Gasselhuber, A. Castan, P. Mayrhofer, C. Grunwald-Gruber, R. Kunert, Bioprocessing of Recombinant CHO-K1, CHO-DG44, and CHO-S: CHO Expression Hosts Favor Either mAb Production or Biomass Synthesis, Biotechnol J 14(3) (2019) e1700686.