A common problem in the development of biopharmaceutical proteins is the lack of understanding around the implications of aggregation, or of their gradual, often partial degradation. Often aggregated or degraded protein variants comprise a small percentage of a biopharmaceutical protein sample, so it is difficult to assess their true impact on product quality and clinical efficacy. The assumption that biopharmaceutical producers adopt is that partially degraded or aggregated protein variants are necessarily undesirable, which may not always be the case. This problem is relevant to the enzyme L-asparaginase from the plant pathogen Erwinia chrysanthemi (ErA, or Erwinase), which is used as a treatment for acute lymphoblastic leukaemia (ALL). This research programme has addressed this issue by examining the effect of potentially adverse changes in ErA relative to its in vitro kinetic properties, and ultimately, to its in vivo pharmaceutical efficacy. Two factors potentially impacting ErA, deamidation and aggregation, were studied in detail. With respect to deamidation, an improvement on published methods for its detection in ErA was developed using a combination of capillary isoelectric focussing (cIEF) and strong chemical denaturants, and this novel method should be widely applicable to other proteins. Deamidation of ErA at selected, theoretically labile residues (N41 and N281, classed as labile using the sequence motif), through expression, purification and characterisation of engineered deamidated mutants, was found to improve, not hinder, the catalytic performance of the enzyme. The structural and kinetic changes imparted by deamidation at these sites were also found to closely mimic an enzyme with high identity and homology (L-asparaginase from Erwinia carotovora). Deamidation of ErA was shown to be at least partially induced during the cell lysis stage of manufacturing, and strategies for minimising creation of deamidated degradants were proposed. ErA aggregation was also studied by examining the aggregate profile in the insoluble, sub-visible 2 - 10µm range. Protein particulates in this size range had been previously postulated as playing a role in immune-mediated allergic reactions during clinical administration of protein biopharmaceuticals such as L-asparaginases. Statistical analyses of allergic response during clinical use of ErA were compared to the lot-to-lot quantification of sub-visible protein particles. The results indicated that ErA allergic response was essentially independent of the level of sub-visible particulates. In summary, the outputs of this research have shown that ErA deamidation and aggregation products are not necessarily deleterious to enzymatic function and clinical efficacy. The work has also demonstrated a set of empirical strategies that may be employed more widely in the development of biopharmaceutical products.