@phdthesis{Bruehlmann2017,
  author    = {Br{\"u}hlmann, David},
  title     = {Tailoring Recombinant Protein Quality by Rational Media Design},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-147345},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2017},
  abstract  = {Nowadays, more than half of the biotherapeutics are produced in mammalian cell lines as a result of correct protein folding and assembly as well as their faculty to bring about a variety of post-translational modifications. The widespread progression of biosimilars has moved the focus in mammalian cell-culture process development. Thereby, the modulation of quality attributes of recombinant therapeutic proteins has increasingly gained importance from early process development stages. Protein quality directly shapes the clinical efficacy and safety in vivo, and therefore, the control of the complex post-translational modifications, such as glycosylation (e.g. high mannose, fucosylation, galactosylation and sialylation), charge variants, aggregates and low-molecular-weight species formation, is pivotal for efficient receptor binding and for triggering the desired immune responses in patients. In the frame of biosimilar development, product quality modulation methods using the potential of the host cell line are particularly sought after to match the quality profile of the targeted reference medicinal product (RMP) as closely as possible. The environment the cell is dwelling in directly influences its metabolism and the resulting quality profile of the expressed protein. Thereby the cell culture medium plays a central role in upstream manufacturing. In this work, concentration adjustment of selected media components and supplementation with a variety of compounds was performed to alter various metabolic pathways, enzyme activities and in some cases the gene expression levels of Chinese Hamster Ovary (CHO) cells in culture. The supplementation of cell culture medium with the trisaccharide raffinose in fed-batch cultures entailed an increase of the abundance of high mannose glycans in two different CHO cell lines. Raffinose especially favored mannose 5 glycans. At the same time, it impaired cell culture performance, induced changes on the intracellular nucleotide levels and even varied the expression levels of glycosylation-related genes. Supplementation with a number of galactosyltransferase inhibiting compounds, in particular fluorinated galactose analogs (alpha- and beta-2F-peracetyl-galactose), consistently decreased the production of galactosylated monoclonal antibodies (mAb). By means of targeted addition during the culture rather than at the beginning, the inhibition was further increased, while limiting detrimental effects on both growth and productivity. High-throughput screening in 96-deepwell plates showed that spermine and L-ornithine also reduced the level of galactosylation. On the other hand, exploratory screening of a variety of potentially disulfide-bridge-reducing agents highlighted that the inherent low-molecular-species level of the proprietary platform cell culture process was likely due to favored reduction. This hypothesis was reinforced by the observation that supplementation of cysteine and N-acetylcysteine promoted fragmentation. Additionally, fragmentation decreased with higher protein expression. At that point, aiming to improve the efficiency in process development, a rational experimental design method was developed to identify and to define the optimal concentration range of quality modulating compounds by calling on a combination of high throughput fed-batch testing and multivariate data analysis. Seventeen medium supplements were tested in five parallel 96-deepwell plate experiments. The selection process of promising modulators for the follow-up experiment in shake tubes consisted in a three-step procedure, including principal component analysis, quantitative evaluation of their performance with respect to the specifications for biosimilarity and selection following a hierarchical order of decisions using a decision tree. The method resulted in a substantial improvement of the targeted glycosylation profile in only two experimental rounds. Subsequent development stages, namely validation and transfer to industrial-scale facilities require tight control of product quality. Accordingly, further mechanistic understanding of the underlying processes was acquired by non-targeted metabolomic profiling of a CHO cell line expressing a mAb cultured in four distinct process formats. Univariate analysis of intra- and extracellular metabolite and temporal glycosylation profiles provided insights in various pathways. The numerous of parameters were the main driver to carry out principal component analysis, and then, using the methodology of partial-least-square (PLS) projection on latent structures, a multivariate model was built to correlate the extracellular data with the distinct glycosylation profiles. The PLS observation model proved to be reliable and showed its great benefit for glycan pattern control in routine manufacturing, especially at large scale. Rather than relying on post-production interpretation of glycosylation results, glycosylation can be predicted in real-time based on the extracellular metabolite levels in the bioreactor. Finally, for the bioactivity assessment of the glycan differences between the biosimilar and the reference medicinal product (RMP), the health agencies may ask for in the drug registration process, extended ranges of glycan variants need to be generated so that the in vitro assays pick up the changes. The developed glycosylation modulator library enabled the generation of extreme glycosylation variants, including high mannose, afucosylated, galactosylated as well as sialic acid species of both a mAb and an antibody fusion molecule with three N-glycosylation sites. Moreover, to create increased variety, enzymatic glycoengineering was explored for galactosylation and sialylation. The glyco variants induced significant responses in the respective in vitro biological activity assays. The data of this work highlight the immense potential of cell culture medium optimization to adjust product quality. Medium and feed supplementation of a variety of compounds resulted in reproducible and important changes of the product quality profile of both mAbs and a fusion antibody. In addition to the intermediate modulation ranges that largely met the requirements for new-biological-entity and biosimilar development, medium supplementation even enabled quick and straightforward generation of extreme glycan variants suitable for biological activity testing.},
  subject      = {Zellkultur},
  language  = {en}
}