University thesis:
Dissertation, Universität Freiburg, 2020
Footnote:
Description:
Abstract: Structural dynamics in proteins forms the basis of phenomena such as vibrational energy transport, protein-peptide interactions and allosteric signal in proteins. To obtain a detailed understanding of these processes, we prepare the systems in a well-defined nonequilibrium state via T-jump or Photo-switching methods. By employing a joint computational-experimental strategy we then follow the relaxation of the systems towards equilibrium. To that end, we first developed and employed efficient simulation protocols that mimic the experimental procedures for the investigation of the aforementioned processes. With regards to vibrational energy transport, extensive nonequilibrium energy transport simulations quantitatively reproduce the experimentally found cooling timesof the proteins at room temperature and predict that the cooling slows by a factor 2 below the glass temperature of water. We further investigate the energy transport pathways and their link with allosteric communication by employing PDZ3 domain as a model system. Using these nonequilibrium simulations, we parameterized the master equation model and determined two scaling rules to predict the rates and pathways of energy transport in general proteins. The master equation model reveals that energy transport is of diffusive nature and the peptide chain along the backbone provides the fastest channel for energy transport with transfer times of 0.5 - 1 ps between adjacent residues. Contact transport, on the other hand, is considerably slower (6 - 30 ps) at room temperature. Furthermore, a quantum correction of factor ≈ 3 of the master equation parameters reproduces the experimental timescales. Second, we have considered Aha-labelled PDZ2 domains and demonstrated that Aha is a minimally invasive, versatile, and sensitive infrared label that can be used to study energy transport, protein-peptide interactions andallostery because the azido (−N3) group of Aha label not only reports on large changesof its chemical environment, but also on small changes of the electronic/electrostatic environment. Third, we have modelled a photoswitchable peptide which is designed in such a way that peptide-protein binding can be controlled by modulating the helical content of peptide. A direct comparison of the simulated α-helical content and experimentally obtained dissociation constants reveals that the binding affinity of S-peptide to S-protein in the Rnase S complex is inversely related to the helicity of S-peptide. Finally, using time-resolved vibrational spectroscopy, nonequilibrium molecular dynamics simulations, and subsequent Markov modeling, we describe the allosteric responseof ligand-switched PDZ2 domain as a change in rugged free energy landscape, witha few structurally well defined states and the dynamics distributed over four decades starting from a nanosecond to microseconds timescales