Year: 2014
Type: dissertation
Abstract: Protein structure, dynamics and function, are inseparable in order to understand the mechanisms of Life at the molecular level. From the structure comes dynamics, and from dynamics many, if not all, protein functions. Usually functional motions operate at timescales and conditions that are far beyond the limits of current experimental techniques. In order to address rationally the complex interplay between these three aspects of proteins, we will deepen into and compare both coarse-grained and atomistic simulation methods. In the first part of this thesis, coarse-grained Elastic Network Models are compared with Molecular Dynamics and experimental flexibility in order to obtain a more accurate representation of protein dynamics. We propose a novel ENM algorithm based on local chain topology, which renders results close to atomistic simulation methods. In the second part of the thesis, the novel method is used to detect dynamical hot spots in the Tyrosine-Kinase HER1, revealing why a number of oncogenic mutations are accumulated at hinge interdomain regions. Near-microsecond long simulations of the detected dynamical mutants are presented which unravel a critical intermediate in the large-scale conformational change that activates the protein. Finally, we explore the validity of the ENM potentials in a number of applications, from sampling transition pathways to understanding how correlated motions in secondary structures transmit information or why dynamics is tightly related with the connectivity of residue networks. The present thesis demonstrates that functional motions are encoded in shape and local-topology, and that perturbations hitting key regions can shift protein dynamics, suggesting a novel oncogenic and evolutive mechanism.
Author: Laura Orellana
Language: English
Subfield: Molecular Biology
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W39664064