Over the last three decades, the Chinese hamster ovary (CHO) cell has progressively established itself as a stable, high yielding mammalian expression system for the production of recombinant proteins – in particular monoclonal antibody (mAb) biotherapeutics. Despite the availability of other platforms, CHO cells continue to dominate as the most utilised system due to their scalability, ability to perform post-translational modifications and long history of regulatory approval. Indeed, over 70% of all therapeutic proteins produced are done so using CHO cells in a highly competitive, and ever-expanding biopharmaceutical market valued in excess of US$180 billion per year.

The production of recombinant biotherapeutics is typically geared towards selecting cells which provide high levels of expression, enabling cost-efficient production of high product-yielding cells. Nonetheless, there remains a proportion of unpredictably low yielding molecules that require extensive, time-consuming process development. These “difficult-to-express” (DTE) therapeutics significantly increase the cost of manufacture and reduce the speed-to-market of innovative treatments relative to “easy-to-express” molecules. Despite their widespread use, little is known about the spatial proteomic landscape of these industrially important cells, which might reveal information about protein location and protein-protein interactions pertinent to mAb secretion and molecular bottlenecks.

This thesis details the cell characterisation of a CHO-K1 production host and two anti-IL-13 recombinant mAb expressing CHO-K1 cells with a 25-fold differential expression yield. This was performed using electron microscopy (Chapter 3) and quantitative tandem mass tag (TMT) labelled proteomics (Chapter 4). Through the optimisation and application of hyperplexed Localisation of Organelle Proteins by Isobaric Tagging (hyperLOPIT); a high-resolution mass spectrometry-based technique, the first, cell-wide proteomic map of the CHO-K1 host is detailed (Chapter 5). This has provided for the unprecedented spatial resolution of thousands of proteins distributed within the CHO-K1 host cell. Furthermore, the hyperLOPIT pipeline was additionally applied to the two recombinant mAb molecules with markedly differing expression characteristics in order to identify changes in subcellular protein localisation (Chapter 6).

Together these studies reveal aberrant ultrastructural morphological differences and protein relocalisations between the DTE and ETE antibody-producing cells, leading to the potential for identification of upstream cell engineering targets to improve production and secretion of high-quality recombinant proteins.