One ring (or two) to hold them all – on the structure and function of protein nanotubes.

Understanding the structural determinants relevant to the formation of supramolecular assemblies of homo-oligomeric proteins is a traditional and central scope of structural biology. The knowledge thus gained is crucial both to infer their physiological function and to exploit their architecture for bionanomaterials design. Protein nanotubes made by one-dimensional arrays of homo-oligomers can be generated by either a commutative mechanism, yielding an 'open' structure (e.g. actin), or a noncommutative mechanism, whereby the final structure is formed by hierarchical self-assembly of intermediate 'closed' structures. Examples of the latter process are poorly described and the rules by which they assemble have not been unequivocally defined. We have collected and investigated examples of homo-oligomeric circular arrangements that form one-dimensional filaments of stacked rings by the noncommutative mechanism in vivo and in vitro. Based on their quaternary structure, circular arrangements of protein subunits can be subdivided into two groups that we term Rings of Dimers (e.g. peroxiredoxin and stable protein 1) and Dimers of Rings (e.g. thermosome/rosettasome), depending on the sub-structures that can be identified within the assembly (and, in some cases, populated in solution under selected experimental conditions). Structural analysis allowed us to identify the determinants by which ring-like molecular chaperones form filamentous-like assemblies and to formulate a novel hypothesis by which nanotube assembly, molecular chaperone activity and macromolecular crowding may be interconnected.