Investigating the Relationship between Chinese Hamster Ovary Cellular Stress Responses, Culture Harvest Day and Formulated mAb Stability

Establishing an appropriate formulation is a crucial step in development of therapeutic monoclonal antibodies (mAbs). The drug product formulation is designed to minimise formation of product instabilities including sub-visible particles (SVPs), aggregates and fragments, and impacts the mode of delivery, dosage, drug format, storage conditions and expiration date. The production of a mAb using Chinese hamster ovary (CHO) cells can be considered as a three phase process; upstream cell culture, downstream purification and formulation/fill-finish activities. Throughout each phase, a therapeutic mAb may be exposed to many stresses, both intracellular during the synthesis of the protein, and extracellular; including pH changes, shear stresses, temperature and concentration changes, all of which can impact on the molecule’s susceptibility to aggregation and SVP formation. Specifically, during culture, cells can be placed under increased endoplasmic reticulum (ER) and oxidative stresses due to cell culture conditions and recombinant protein loads on the cell, which may result in compromised mAb yield and quality. Although much work has been undertaken to investigate how such intracellular stresses impact mAb titres and quality, there are no published studies investigating how cellular stresses change across culture and how such stress may impact formulated mAb stability. The work presented in this thesis explores the relationship between intracellular stress, cell culture processes, culture harvest day and formulated mAb stability of three model therapeutic mAbs (denoted 109, 4212 and 184) by profiling ER stress throughout culture at the transcript and protein level, and by comparing the stability profile of purified mAb material from an ‘early’ and ‘late’ harvest of 400 mL roller bottle cultures (day 9 and 13 respectively) and 10 L disposable bioreactors (days 8 and 13 respectively). Furthermore, the use of ER and oxidative stress reporter constructs to monitor these stresses in real time during mAb production was also investigated.

Overall, harvest day was shown to impact mAb stability, as assessed by the propensity of formulated material to form SVPs. Interestingly, the intracellular ratio of mAb heavy chain (HC) to light chain (LC) mRNA also impacted on the stability of harvested mAb material from 400 mL cultures when formulated in different buffer compositions. Furthermore, studies on mAb material generated using 400 mL roller bottle cultures also showed inherent molecular differences between cell lines cultured under fed-batch and batch conditions, where cell lines 4212 and 184 displayed similar titre, growth and biomarker profiles under batch culture; but under fed-batch behaved differently. Stability studies of both 400 mL roller bottle and 10 L disposable bioreactor material showed a cell line and/or mAb and/or culture process specific relationship between harvest day and SVP formation. Formulated 4212 mAb material harvested on day 9 of roller bottle culture produced fewer SVPs than that from day 13 indicating a relationship between day of harvest and mAb stability, whereas mAb 184 showed no harvest day trend. When cultured in 10 L disposable bioreactors, however, cell line/mAb 184 showed a strong harvest day trend, with mAb 184 material from day 8 of culture producing fewer SVPs and appearing more opalescent than that from day 13. Furthermore, CD spectroscopy highlighted conformational differences between mAb 184 material from the two harvest days, withRNAseq revealing differences between cellular transcript profiles on days 6, 8 and 13 of culture for cell line 184. Finally, reporter constructs to monitor ER and oxidative stress during fed-batch culture showed varying levels of sensitivity. During validation studies, the ER stress response element (ERSE) proved unresponsive to chemical induction of ER stress, however, the antioxidant response element (ARE) responded to chemically induced oxidative stress across a range of drug concentrations, however there was insufficient oxidative stress perceived by cells during culture to activate the reporter during mAb synthesis. Collectively the data presented in this thesis provides novel insights into the impact of bioprocessing on formulated mAb stability, and shows how harvest day can impact on the stability of formulated mAbs and demonstrates that the relationship between upstream cell culture and the stability of an expressed therapeutic is complex and dynamic.