Selection and evolution of aggregation resistant proteins

Over the last 40 years, proteins have emerged as highly effective therapeutics due to their endogenous specificity. While the structural and biophysical properties of protein scaffolds allow the formation of highly avid complexes, the inherent metastability of proteins can result in local or global unfolding that can lead to inactivation and/or protein aggregation. As a result, the production and formulation of biopharmaceuticals can be hindered by protein aggregation which can occur at every stage of the manufacturing process; ultimately jeopardising the successful development of promising candidates from becoming the next blockbuster biologic. Aggregation compromises the quality, stability, and safety of a drug product, yet the ability to identify ‘manufacturable’ candidates with long-term stability during lead isolation and optimisation remains challenging. Similarly, the ability to predict the aggregation propensity of proteins associated with protein aggregation diseases is also arduous, and much remains unknown about the fundamental processes driving protein aggregation in these diseases. There is thus an important and currently unmet need to be able to identify protein sequences that may have undesired properties and to engineer their sequences to improve their properties. Investigating protein aggregation and stability can be laborious, due to the difficulties in expression and purification for in vitro analysis and since aggregation can occur through a variety of mechanisms. The work presented in this thesis employs a tripartite β-lactamase platform to characterise the aggregation propensity of biopharmaceuticals that circumvents the need for recombinant expression and downstream analysis. This system can distinguish between aggregation and non- aggregation prone sequences, offering a powerful tool for assessing protein aggregation and stability earlier in the industrial pipeline. This study also developed a directed evolution methodology that can be used as a novel strategy to modulate the aggregation propensity of protein therapeutics, to evolve ‘manufacturable’ biopharmaceuticals early during industrial development. Importantly, the approach does not require any structural knowledge or prior biophysical information about the protein of interest. Finally, the application of this platform to disease-related proteins enabled the identification of hotpot residues that differ between germline and patient sequences in light chain amyloidosis, that may further the understanding of the processes that underpin aggregation diseases. Overall, this platform provides a new approach for the rapid identification of aggregation resistant proteins and to provide insight into the molecular mechanism of aggregation