Deciphering conformational ensembles and communication pathways in biomolecules
Time: Fri 2020-12-18 09.00
Location: Zoom livestream https://kth-se.zoom.us/j/61191023989, (English)
Subject area: Biological Physics
Doctoral student: Annie M. Westerlund , Biofysik, Science for Life Laboratory, SciLifeLab
Opponent: Dr Kresten Lindorff-Larsen, University of Copenhagen, Köpenhamn Danmark
Supervisor: Universitetslektor Lucie Delemotte, Biofysik, Science for Life Laboratory, SciLifeLab; Professpr Erik Lindahl, SeRC - Swedish e-Science Research Centre, Fysik, Science for Life Laboratory, SciLifeLab, Biofysik
Life is constructed by small building blocks called cells. Proteins are the biomolecules within these cells which carry out different functions. Proteins located in the cell membrane, for example, allow for the transport of ions and small molecules, as well as communication, across the cell membrane. These events control important processes such as the heart beat, muscle contractions and immune system regulation. Membrane proteins are therefore often targets of new drugs.
Molecular dynamics simulations provide an atomistic view of biomolecule movements. By simulating proteins in environments mimicking their native ones, we are granted access to the molecular mechanisms which govern biologically important events. This includes transitions between structural states, ligand binding or communication between protein domains via allostery. Knowing the atomistic details behind such mechanisms is fundamental to novel drug discovery. Historically, the computationally intensive calculations associated with molecular dynamics have limited the accessible timescales. The rapid developments in both software and hardware, however, have facilitated longer simulations of larger systems, leaving researchers with the task of deciphering these large datasets. The topic of this thesis covers development and application of such data-driven analysis to understand molecular simulations of proteins.
The papers are divided into two parts. The work associated with Part 1 (Papers I-IV) considers estimation of free energy landscapes, extraction of metastable states, and inference of important structural features related to these states. Here, the tools are specifically used to study the Ca2+-dependent state ensembles of the regulatory protein calmodulin. The papers of part 2 (Papers V-VI) instead consider the use of network analysis to study protein allostery in membrane proteins. This is used to reveal the coupling between the voltage sensor domains and the pore of the voltage-gated potassium channel KCNQ1. It is also used to investigate how lipids and small molecules may modulate allostery in membrane proteins. In summary, the work presented here uncovers mechanistic details of physiologically crucial proteins, and serves as a stepping stone towards data-driven biophysical research and drug discovery.