Preview

PDF, English Download (15MB) | Terms of use

Abstract

Nature has come up with an enormous variety of protein three-dimensional structures, each of which is thought to be optimized for its specific function. A fundamental biological endeavor is to uncover the evolutionary driving forces for discovering and optimizing new folds. A long-standing hypothesis is that fold evolution obeys constraints. Aiming at elucidating those constraints, we evaluated some physical quantities for a large number of biological molecules. Firstly, flexibility was estimated via two independent methods: CONCOORD, which predicts conformational ensembles for atomic protein structures using geometrical constraints, and elastic network models, a simple coarse-grain model. Foldability was measured by Contact Order, which can predict the folding rate of a protein by measuring the distance between native contacts within the protein. Lastly, mechanical strength was predicted with Langevin Dynamics simulations of the conventional Go-type models of proteins, a coarse-grained model based on the X-ray structure, under force. We mapped those physical quantities onto a phylogenomic tree of protein structures resulting from the analysis of the abundance of ~3,000 protein families. Bimodal trends were observed for the different physical quantities suggesting a turnover at around ~1.5 billions years ago. This turnover corresponds to the apparition of multicellular organism that could have drastically modified the constraints applied on the evolution of protein structures. More specifically, before ~1.5 Gya, we observed an increase of foldability and a decrease of mechanical stability that might be the result of a concerted need for fast folders and compact proteins resulting from molecular compartimentalization, i.e. the rise of cells. On the contrary, after ~1.5 Gya, we observed a decrease of foldability and an increase of mechanical stability that suggest a need for mechanical stability probably related to the rise of multicellular organisms with increased mechanical stresses between cells. The loss in foldability after the big bang might be due to that cells started to make use of proteins such as chaperones or other advanced mechanisms thereby removing, at least partly, the constraint for fast folders. Taken together, we identified physical constraints that are likely to play a role in the evolution of protein structures. Our global approach opens avenues for a more comprehensive analysis of genomic and structural data available. Improving our view on protein structure evolution is likely to bring more insights into their functioning. Additionally, it could help constructing a network based view of protein structures evolution improving the classification of the known protein catalogue and aiding the design of new protein structures.