Computational prediction and redesign of aberrant protein oligomerization

Non-native intermolecular contacts often lead to aberrant protein oligomerization and aggregation. This phenomenon is behind the onset of several human disorders and is a bottleneck for the production of proteins of biotechnological interest. Intrinsically disordered proteins have evolved to avoid aberrant oligomerization, but mutations or aging-promoted degeneration of the protein quality machinery might result in their aggregation. Folded globular proteins are not completely protected from aggregation, mostly because the physicochemical properties stabilizing their tertiary and/or quaternary structures are very similar to those leading to non-native oligomerization. Once these properties are known, it becomes feasible to predict the aggregation propensities of proteins and to design them to disfavor aggregation-prone contacts. In this chapter, we describe how computational approaches can assist the identification of the aggregation-prone sequential or structural regions leading to aberrant oligomerization and how these tools can be employed to predict pathogenic mutations or to design biotherapeutics with optimized solubility.