Experimental and Computational Approaches to Control Mammalian Cell Proliferation, Protein Production, and Post-translational Modifications

Therapeutic antibodies, with their high specificity and flexibility, are responsible for treating many acute and chronic diseases. In particular, monoclonal antibodies remain the largest sector in sales and market growth in the biotechnology industry. The production of antibody biologics involves a rigorous pipeline of target identification, cell line development, upstream processing, downstream processing, and product formulation - each facing its unique challenges.

Specifically, in the upstream sector, two main characteristics critically define the production cost and quality of the antibodies made: antibody titer, or the total amount of protein harvestable, and protein critical quality attributes (CQAs). The first determines the batch consistency and cost of delivery, while the latter determines the safety and efficacy of the antibody product. In this thesis, we used a small molecule, rosmarinic acid, which significantly improved cellular proliferation as a model to study pathway stimulations to enhance antibody production. In addition, through a combination of media supplementation and process variation strategies, we improved antibody production maximally threefold without compromising either cellular-specific productivity or glycosylation attributes.

In addition, to meet market demands and deliver safe and potent therapeutics in time, both high production and quality standards must be met. This requires parallel screening and development of clonal cells and optimal cell growth conditions at the bench scale before manufacturing. The complexity of cellular pathways and the different post-translational needs of products have created significant challenges for rational process design strategies. Importantly, with a rising interest in developing redox and endoplasmic reticulum (ER) stress biosensors to unravel and control cellular behavior for protein production runs, reports of varied cellular stress effects on cell growth and protein production convolute understanding; and determination of the effect of cellular stress on antibody quality is still deficient. In the second part of the thesis, we address this unique yet complicated problem by identifying representative oxidative and ER stressors combined with design-of-experiment (DOE)-guided experimentation to implement a high-dimensional space accommodating different degrees of cellular stress to study their impacts on the cell culture process and protein quality. We also develop a high- throughput assay platform to detect both absolute and relative glycosylation simultaneously and examine how these attributes may be controlled via hybrid modeling to interface with rapid digitization in the bioprocessing industry. Further, in the third part of my thesis, I use small molecules and empirical models to actively control antibody glycosylation, one of the most important CQAs, and discuss how the implementation of process analytical tools for glycosylation may benefit the overall control process.

In the last part of this thesis, we explore protein quality control in the ER and nucleus in a protein misfolding disease driven by tau protein. Tau protein aggregation drives the onset and progression of Alzheimer’s disease, the 6th leading cause of death in the United States reported in 2021. While tau protein commonly localizes in the cytosol, under pathological conditions it has been reported to associate with the nucleus and endoplasmic reticulum. Here, we use protein localization motifs to reroute tau localization to study its expression and phosphorylation characteristics in the ER and nucleus. Finally, possible mechanisms driven by the tau variant P301L, a common mutation in Alzheimer’s disease, will be discussed through an omics analysis of miRNA sequencing from exosomes derived from wildtype and P301L tau variants in neuronal cells.