Modeling of proteins with low-resolution experimental data

Dnr:

SNIC 2016/34-40

Type:

SNAC Large

Principal Investigator:

Erik Lindahl

Affiliation:

SciLifeLab, Stockholms universitet

Start Date:

2017-01-01

End Date:

2018-01-01

Primary Classification:

10603: Biofysik

Secondary Classification:

10602: Biokemi och molekylärbiologi

Tertiary Classification:

10407: Teoretisk kemi

Webpage:

http://www.lindahllab.org/

Allocation

Abstract

The goal of our work is to understand how structure, flexibility, and conformational transitions in proteins explain biological function. We use both simplified models and molecular simulations in combination with experimental investigations by us and other teams, in particular electrophysiology and cryo-EM. This approach makes it possible to study features of proteins that are not readily available in experiments, such as transitions between states (particularly important for membrane proteins) and finding new ways to use low-resolution experimental data. We are particularly interested in ligand-gated ion channels such as the GABA receptors and prokaryotic relatives (GLIC), where our goal is to understand the transitions between open, closed, and desensitized states, and how small drugs allosterically modulate this process. We increasingly believe the transitions happen asymmetrically, and would like to use simulations to study how the free energy of transitions depend on whether the subunits move synchronously or not - this will be based on transition pathways we recently obtained from Langevin dynamics of the principal components of experimental structures. We also propose to continue our investigations of the Shaker voltage-gated channel. We recently published how poly-unsaturated fatty acids interact with the voltage-sensors, and would now like to use simulations to understand how small molecules that mimic these fatty acids interact with the channels to help identify binding mechanisms and iterate with experimental studies to design drug candidates. Cryo-EM is becoming a new important research direction for us, and we have spent significant efforts on developing new GPU-accelerated methods for density reconstruction from micrographs in the RELION code. We want to combine cryo-EM data with simulations, and will use a new method to apply Bayesian restraints to several proteins (in particular ribosomes and ion channels) based on the cryo-EM electron density. Many of these structures are not well represented with a single all-atom model in blurred regions with large conformational flexibility or low resolution, but with our approach we provide an ensemble of conformations most likely to have generated the experimental data, instead of a single model. We would also like to investigate the possibility of using this technique to explore targeted pathways between states, based on experimental structural data. This approach is not limited to cryo-EM, but equally useful for other low-resolution methods such as SAXS/SANS - and ideally we should be able to use the Bayesian formalism to correctly apply multiple different types of data in simulations, even when the measurements might not be entirely consistent (e.g. different states). Finally, we are continuously developing GROMACS, with the latest version released this summer; this has turned into one of the most widely used codes in the world for molecular simulation, and the SNIC resources are critical to help us tune performance, parallelization and scaling for state-of-the-art machines such as the Cray XC40 at PDC, not to mention that we are developing a number of algorithms such as efficient lattice summation methods for Lennard-Jones interactions.