Conformational stability as a design target to control protein aggregation

Non-native protein aggregation is a prevalent problem occurring in many biotechnological manufacturing processes and can compromise the biological activity of the target molecule or induce an undesired immune response. Additionally, some non-native aggregation mechanisms lead to amyloid fibril formation, which can be associated with debilitating diseases. For natively folded proteins, partial or complete unfolding is often required to populate aggregation-prone conformational states, and therefore one proposed strategy to mitigate aggregation is to increase the free energy for unfolding (Delta G(unf)) prior to aggregation. A computational design approach was tested using human gamma D crystallin (gamma D-crys) as a model multi-domain protein. Two mutational strategies were tested for their ability to reduce/increase aggregation rates by increasing/decreasing Delta G(unf): stabilizing the less stable domain and stabilizing the domain-domain interface. The computational protein design algorithm, RosettaDesign, was implemented to identify point variants. The results showed that although the predicted free energies were only weakly correlated with the experimental Delta G(unf) values, increased/decreased aggregation rates for gamma D-crys correlated reasonably well with decreases/increases in experimental Delta G(unf), illustrating improved conformational stability as a possible design target to mitigate aggregation. However, the results also illustrate that conformational stability is not the sole design factor controlling aggregation rates of natively folded proteins.