Multi-omics analysis of Chinese hamster ovary cell-based biopharmaceutical production processes

Abstract

Monoclonal antibodies (mAbs) are an important group of biotherapeutics treating inflammatory, neurological, and oncological diseases. For more than three decades, the bioprocessing industry has used Chinese hamster ovary (CHO) cells, primarily in fed batch bioprocesses, to produce therapeutic recombinant proteins such as mAbs. Today, there is continual demand to maximize CHO cell culture productivity while maintaining product quality in order to reduce the cost of mAb production. Advancements in liquid chromatography coupled to mass spectrometry (LC-MS) have enabled the rapid profiling of many cellular proteins and metabolites with high sensitivity while simultaneously minimizing required sample amounts. This has paved the way for researchers to conduct 'omics experiments with the objective of better understanding the CHO cell biology and characterization of the relationship between bioprocess conditions and phenotypes. Multi-omics, i.e. the use of transcriptomics, proteomics, metabolomics and lipidomics together in a systems biology approach, is an emerging field gaining momentum in bioprocess development. Through the combined information from multiple 'omics experiments, researchers are able to gain a more comprehensive picture of the biochemical changes taking place during bioprocessing. This dissertation begins with an introductory chapter that provides an overview of bioprocessing, mammalian cell metabolism, and advancements in LC-MS technologies. A background of each of the 'omics approaches conducted in this dissertation is provided while a brief summary of how multi-omics have been previously used in the biotechnology industry is described. Chapter 2 focuses on lipid metabolism during the complete range of mAb bioprocessing using CHO cells. In this chapter, an LC-MS method is developed, taking advantage of both positive and negative mode electrospray ionization (ESI) for comprehensive lipid profiling. Novel bioinformatics techniques are used to investigate the changes in lipids inside the cell over the course of the bioreactor cultivation. The results revealed the importance of the CHO lipidome on cellular growth and specific productivity. The lipidomics results were combined with transcriptomics (RNA sequencing) and polar metabolite data, in a pathway analysis approach, to highlight the role of lipid metabolism in cell growth and productivity. In Chapter 3, we conducted a multi-omics experiment exploring the metabolic changes that resulted in reduced titer and increased cell death when insufficient levels of the amino acid cysteine (Cys) were fed in a fed batch process. Intracellular and extracellular samples were analyzed throughout the time course of the bioreactor cultivation. The analysis demonstrated low expression of Cys biosynthesis enzymes led to depletion of the amino acid and resulted in a significant decrease in cellular antioxidants such as glutathione, hypotaurine and taurine. The reduction had detrimental effect on the cellular redox balance, leading to apoptosis. By using pathway analysis, key metabolic enzymes and metabolites were identified for bioprocess monitoring. This work illustrates the importance of redox balance to cellular health and bioprocess production. Finally, in Chapter 4 the findings from Chapter 3 are used and expanded upon to understand in depth the CHO cellular response to redox imbalance as well as amino acid (Cys) depletion. In this chapter, we delve further into the impact of insufficient Cys feeding on the bioprocess, focusing particularly on the negative effects on specific productivity and energy metabolism. Through multi-omics analysis, we found increased endoplasmic reticulum (ER) stress when Cys levels were depleted leading to mitochondrial stress and dysfunction. Together, ER stress and Cys depletion activated the amino acid response which led to an undesirable shift in energy metabolism (Krebs Cycle) and a reduction in specific productivity. An epilogue is provided to discuss future studies and prospects. By leveraging the unique dual role of Cys in both redox balance and amino acid metabolism, multi-omics analysis was able to reveal new opportunities for bioprocess improvement while suggesting relevant and diverse sets of intracellular and extracellular bioprocess markers. The findings described in this dissertation demonstrate the power of multi-omics analysis in biotherapeutic process development and manufacturing.