From Static Structures to Free Energy Landscapes: Characterizing Conformational Transitions in Biological Macromolecules
Time: Tue 2023-06-13 09.00
Location: Air&Fire, Tomtebodavägen 23, Solna
Language: English
Doctoral student: Cathrine Bergh , Biofysik
Opponent: Professor Gerhard Hummer, Max Planck Institute of Biophysics
Supervisor: Professor Erik Lindahl, Biofysik
QC 2023-05-17
Abstract
To be alive means to temporarily counteract the fundamental dispersive driving force described by the second law of thermodynamics, eventually leading all systems to decay and disorder. In cells, this task is partly carried out by proteins - small specialized molecular machines that utilize free energy to maintain the proper functioning of the cell. Malfunction of a protein, often caused by genetic mutations, can lead to death and disease, underscoring the importance of understanding their function in order to develop new drugs and therapies. The function of a protein can in principle be fully described by its free energy landscape, a probability distribution that maps the relationship between protein structure and function through motion. However, constructing such free energy landscapes through experimental techniques alone is close to impossible. Some techniques can generate static snapshots of protein structures but do not yield direct information about the function, while other methods capture functional aspects but do not have the resolution to uncover the structural changes involved. Computer simulations in the form of molecular dynamics simulations can bridge this gap but are associated with limitations in the form of sampling relevant timescales and distilling high-dimensional data into a form that properly describes protein function.
The aim of this thesis is to develop and apply methods that can overcome the sampling and analysis problems associated with molecular dynamics simulations, assess their predictive power through experimental validation, and explore how such models can be effectively combined with various experimental techniques, such as SANS, cryo-EM, and electrophysiology. Finally, these methods are applied to reveal multiple new structural and functional aspects of pentameric ligand-gated ion channels (pLGICs), which are of great importance to synaptic signal transmission between nerve cells.
The work presented in this thesis can be characterized according to three different themes. First, I describe the development of a coarse-grained simulation method that generates an approximate pathway between two known states, and how this can be used to achieve better sampling in molecular dynamics simulation. Then, I investigate how enhanced sampling and Markov state modeling can be used to estimate free energy landscapes of pLGICs, and how such methods can work alongside and inform experimental methods to reveal many new aspects of pLGIC structure and function. Finally, I explore the use of Markov state modeling in the drug discovery field. This thesis thus contributes to computational method development and the understanding of pLGIC function - two different aspects with potential to contribute to the development of new drugs.