Determination of Bioprocessing Variables that Influence Particle and Aggregation Formation of Therapeutic IgGs

One of the dominant classes of protein based drugs, which continues to contribute to the growth and success of biologics, are recombinant monoclonal antibody (mAb) therapies. In preparing recombinant antibodies for biotherapeutic applications, the stability and maintenance of this is of utmost importance; not just for the activity of the molecule but also in terms of the safety of the drug and this must be demonstrated to regulatory authorities before a biotherapeutic receives approval for use. One such measure of protein integrity, quality or stability is the assessment of soluble aggregates and sub-visible particles present in formulations after stressed stability studies. The presence of such particles and aggregates in recombinant biotheraputics is potentially of concern in the manufacture of such biologics due to the reported immunogenic responses observed in patients when these biologics are administered; which is believed to be associated with the presence of aggregated antibody species. In addition to this, the in vivo efficacy of the biologic is inevitably compromised upon formation of such species. As such the regulatory bodies require that aggregate and particle levels are monitored and reported in the specification of a biologic.

The manufacturing of recombinant mAbs using mammalian expression systems involves three general stages; upstream, where the target molecule is cloned and expressed in the system of choice; downstream, where the recombinant protein material is recovered from the fermentation supernatant and formulation whereby the molecule is ‘formulated’ to maintain its stability and for delivery to patients. Throughout this process the mAb molecules experience a wide variety of stresses such as temperature, high concentrations, interactions with host cell impurities, varying pH and shear stresses that could all potentially influence a molecules susceptibility to aggregate or form particles. Whilst there has been much research at the amino acid level to determine how primary sequence influences propensity of molecules to aggregate, there has been little work investigating how the whole manufacturing process might influence the susceptibility of mAbs to aggregate and form particles. The focus of this work was therefore to investigate whether three variables during manufacturing of mAbs; feeding during fermentation, harvest day and the exclusion/inclusion of an additional wash step during purification, influence the levels of aggregates and particles in antibody formulations post-temperature stressing. Atomic force microscopy was also used to investigate the mechanism of mAb particle formation and aggregation using a model system under forced aggregation conditions.

The results from these studies confirmed that washing of the protein A column during mAb purification can influence subsequent particle formation. The inclusion of a specific wash step during protein A purification was effective at lowering particle levels in formulations of a model mAb compared to those samples which were purified without. The wash step was shown to remove mAb fragments and molecular chaperones and other host cell proteins from the Chinese hamster ovary cell host. The re-introduction of these species back into mAb solutions resulted in a subsequent increase in particle formation suggesting that species present in the material removed by the wash fraction influenced the particle numbers observed. With regard to culture conditions and feeding of the cells, batch cultures produced a less stable molecule compared to the fed-batch cultures as determined by the propensity to form particles and this could be related to the intracellular stress being experienced by cells in during culture. Cell lysis is influenced by the intracellular stress levels of the cultures, which is in-turn influenced by the feeding regime and harvest day. The levels of mAb fragments and sub-visible particles increased in fed-batch and batch Mab-184 samples, as culture viability decreases; which was mirrored by changes in the expression of certain stress inducible genes. Purification of the samples with the wash step reduces the particle levels to similar levels across all harvest days suggesting that the wash step is not only removing misfolded protein released into the culture supernatant upon cell lysis, but also removes species that are present in the cell culture supernatant which contribute to antibody particle formation. These results highlight the need for better monitoring of intracellular stress levels during CHO culture for informing harvest decisions as the intracellular state of the cells influence the quality of the final therapeutic product. Finally, using AFM particle formation of a model mAb is characteristic of a coagulation mechanism after one hour of temperature stress. Collectively together these studies demonstrate the importance of both upstream and downstream processing decisions on the subsequent propensity of mAbs to form particles and hence the need to consider the whole bioprocess to obtain the best possible product with regard to aggregate and particle amounts.