| Abstract
|
:
|
Proteins need to react quickly to constantly changing environments. Their ability to evolve new structures and functions is manifested in many of their physical and chemical traits. One such trait is the modular structure of proteins, which enables rapid rearrangements for the creation of new scaffolds mediating new functions. The advantage of modularity lies in the use of existing "off-the-shelf" elements as building blocks for new proteins rather than building new ones from scratch. The modular organization of proteins may also refer to the evolutionary origins of the protein world. Proteins with high internal symmetry, for example, are thought to have evolved from short gene segments encoding rudimentary structural modules that assembled noncovalently, and thus yielded functional oligomeric proteins. At later stages, these gene segments were duplicated, fused, and rearranged, to yield the single polypeptide chain proteins we recognize today. The hypothesis of oligomeric precursors has been inspired by the dominance of folds such as β/α-(TIM)-barrels, or β-propellers, that are comprised of repetitive modules. Albeit, there is very little experimental evidence to support this hypothesis, or indeed, to show that the apparent modularity of contemporary β/α-(TIM)-barrels, or β-propellers, may serve as an evolutionary starting point and yield new tertiary, or quaternary, structures. In this study, we aimed at reconstructing the evolutionary scenarios leading to the emergence of symmetrical, modular folds, and isolate evolutionary intermediates that correspond to their putative oligomeric precursors. Our model case has been tachylecin-2—a 236-amino acid protein that adopts a 5-bladed β-propeller fold with five sugar-binding sites. By means of laboratory evolution, we have identified putative evolutionary routes leading to functional lectins by covalent (duplication and fusion), and non-covalent (oligomerization) assembly of various elements taken from tachylectin-2. We describe the duplication and fusion of five modules (or sequence repeats) grafted from tachylectin-2's sequence to yield functional lectins. Subsequently, ∼100-amino acid segments truncated from the full-length lectin were isolated, that spontaneously assemble to give functional homo-pentamers with approximately twice the molecular weight of wild-type tachylectin-2. The 3D structure of this newly evolved oligomeric lectin was solved. The homo-pentamer assembly is enabled by new strand exchanges that connect between the subunits in a way similar to the "Velcro" closure of tachylectin-2, but involving novel inter-strand interactions. Intriguingly, the subunits (that are identical in sequence) adopt two different structures within this pentamer, and thus enable the assembly of two 5-bladed β-propellers connected by a linker. The structure therefore highlights the key role of symmetry and modularity in the assembly of β-propellers, and β-strand based proteins in general, and of conformational diversity (or protein metamorphism) as a key component in evolutionary innovation. Following the observation that many lectins, especially in plants, make use of structural reorganization, and changes in quaternary structure in particular, to gain new ligand binding specificities, we are using the new oligomeric lectin to follow how their modular nature, and their oligomeric arrangement, may accelerate the evolution toward new glycosylated protein targets. To this end, we have evolved the pentameric lectin to recognize glycoproteins that are not bound by the monomeric, wild type tachylectin-2. The newly evolved specificities were monitored by high-throughput analysis of the evolved lectins, to reveal a dramatic change in their binding patterns, including the recognition of new sugars. Together, these results highlight the important role of oligomeric assemblies in the evolution of new functions and folds. Finally, to further establish the role of modularity in protein evolution, we attempted to use a highly conserved structural motif as the basis or new functional proteins. By means of non-homologous recombination, we recombined highly conserved motif taken from the active site of serine hydrolases, with short, random protein fragments, and selected the resultant library for functional proteins.
|