Department of Theoretical and Computational Biophysics, Max-Planck-Institute for Biophysical Chemistry, Göttingen, Germany
Local Project ID:
HPC Platform used:
SuperMUC of LRZ
Ion channels play a fundamental role in maintaining vital electrochemical gradients across the cell membrane and in enabling electrical signaling across cells. Key characteristics of ion channel function that can be experimentally quantified include ion permeation rates and selectivities. In this project, the functional mechanism of a very important class of ion channels is investigated with the help of molecular dynamics simulations. The computer simulations exhibit a wide range of GLIC states from completely closed to wide open, with conductance and selectivity for the open state in agreement with experimental values. The scientists are now beginning to investigate the intricate opening/closing mechanism in detail to ultimately explain it from a physics perspective.
All living organisms are made of cells. Cells separate their interior from the exterior environment by the cell wall, the so-called plasma membrane. Embedded in this membrane are (among many other functional proteins) various channel proteins that control what goes in and out of the cell. A channel protein acts as gate and gatekeeper rolled into one: Depending on its type, it will only let specific molecules pass when it is open. The so-called aquaporins, for example, only let water molecules pass. Another important type of channels control the influx and efflux of ions and are therefore called ion channels. These are fundamental to all living beings as they maintain vital electrochemical gradients across the cell membrane and enable electrical signalling across cells. Key characteristics of ion channel function that can be measured experimentally are ion permeation rates and selectivities, i.e. preferences for types of ions.
A particular group of ion channels with very interesting properties are the pentameric ligandgated ion channels (pLGICs). They play a key role in fast synaptic signal transduction in brain and muscle and are, as the name suggests, composed of five similar (or identical) subunits. As illustrated in Fig. 1 (left), they consist of a large extracellular domain (ECD) and a somewhat smaller transmembrane domain (TMD), which includes the channel’s pore. While pLGICs remain closed in their resting state, they can be activated by binding of small molecules (ligands) to the ECD. An intriguing property of pLGICs is that ligand binding at the exterior side leads to structural rearrangements that propagate through the channel to the remote TMD, where they trigger pore opening (Fig. 1). This strong coupling between ligand binding and pore opening seems to be the universal gating mechanism in all pLGICs.
In this project, the scientists aim to shed light on this intricate gating mechanism by characterizing and explaining it in physical and energetic terms. To that end, an all-atom model of a pentameric ligand-gated ion channel was build from high-resolution atomic structures that have recently become available. To mimic its natural cellular environment, the channel was embedded in a lipid membrane and surrounded by water and ions.
To study the gating process of pentameric ligand-gated ion channels in general, the GLIC channel from Gloeobacter violaceus is used as a prototypic system. GLIC is structurally highly similar to physiologically important pLGICs in humans, like acetylcholine, GABA, and glycine receptors. In contrast to other pLGICs, GLIC is proton regulated, which means that an extracellular pH drop is the trigger for pore opening (Fig. 1).
Well-resolved X-ray crystal structures exist for what is proposed to be the open (PDB identifier: 4HFI ) and the closed state (PDB: 4NPQ ) of GLIC, which are used as starting points for all-atom molecular dynamics (MD) simulations. As the time-scale of GLIC opening (from extracellular pH drop until pore opening) is about a millisecond, it is not possible to trigger and directly observe the gating process in a single, long MD simulation. For computational reasons, MD trajectory lengths are currently limited to about a microsecond. Therefore, the scientists took an alternative approach by simulating the individual stages of the gating process in a large ensemble of simulations. Using the two all-atom models built from the 4HFI and 4NPQ structures, about 50 individual stages were linearly interpolated between those with an increasing degree of openness.
The fundamental question addressed in the initial phase of the project is whether the 4HFI and 4NPQ experimental structures indeed capture the conducting and the non-conducting state of GLIC. Confirming this is a prerequisite for further investigations into GLIC’s opening/closing mechanism.
To determine the conductance properties of the 4HFI and 4NPQ experimental structures as well as of the in-between interpolated stages, computational electrophysiology (CompEL) simulations [1,6] were performed with the GROMACS molecular dynamics package [2,3]. In a CompEL setup, two MD systems are stacked on top of each other, with each system consisting of a channel in a membrane surrounded by water and ions (Fig. 2). This way, in periodic boundary conditions, two separate compartments are formed such that ions can get from one compartment to the other only by passing the channel. A charge imbalance is applied between the compartments by placing a few more positive ions in one compartment and a few more negative in the other. This leads to a potential difference ΔU which induces an ionic current through the channels, if they are open and conducting. To prevent that the current dissipates the ionic charge imbalance, ion/water pairs are artificially exchanged between the compartments as needed to restore the original charge imbalance, leading to a steady flux of ions through the channel(s). This protocol allows to determine the conductance properties of a channel like in a real electrophysiological experiment.
The questions addressed with the simulations are: (i) Is the 4HFI crystal structure indeed conducting and the 4NPQ structure indeed nonconducting? (ii) What happens during pore opening? What distinguishes the conducting from the nonconducting structural state? (iii) Is the conductance behavior determined by the transmembrane part of the channel alone or does the extracellular domain also play a role?
The time-scale of GLIC opening of about a millisecond renders the direct simulation of this process impossible, as trajectory lengths are currently limited to about a microsecond due to computational reasons. This, together with the fact that the pLGICs are rather large proteins (the complete simulation system comprises about 600,000 particles), makes this project extraordinarily challenging and only feasible on HPC resources like SuperMUC.
To address the above mentioned questions, the scientists have determined the ionic conductivity of the GLIC channel for 50 stages along the opening coordinate in CompEL setups. MD simulations were carried out, both for the whole channel as well as just for the transmembrane part, to study the effect of the extracellular domain.
The simulations clearly established that the 4NPQ structure is nonconducting and that the 4HFI structure is conducting sodium ions (top panel of Fig. 3). The 4HFI conductivity calculated from the simulations lies in the low pS range and is thus compatible with experimental results (yellow region in Fig. 3).
Moving from left to right in Fig. 3 is equivalent to moving through the stages the channels visits in its opening motion. During the transition from closed to open, conductivity sets in at the same time as the pore fills with water (see left scale of Fig. 3). The more water is in the pore, the higher the conductivity gets.
The scientists could also determine that the transmembrane part of the GLIC channel alone (without the large extracellular domain attached) shows a similar conductance behavior, however the conductivities are generally smaller and set in later in the opening process (bottom panel of Fig. 3). This indicates that the extracellular part also modifies the conductance behaviour of the pore.
Although the simulations themselves are now successfully completed, the scientists are still a long way from concluding their research. The data generated by the simulations still needs to be thoroughly evaluated and interpreted. It forms the basis for further research into the molecular mechanism and the gating behaviour of GLIC and pentameric ligand-gated ion channels in general.
1. Kutzner, C.; Köpfer, D.; Machtens, J. P.; de Groot, B. L.; Song, C.; Zachariae, U.: Insights into the function of ion channels by computational electrophysiology simulations. Biochimica et Biophysica Acta – Biomembranes 1858 (7, B), pp. 1741 - 1752 (2016)
2. Páll, S.; Abraham, M. J.; Kutzner, C.; Hess, B.; Lindahl, E.: Tackling exascale software challenges in molecular dynamics simulations with GROMACS. In: Solving Software Challenges for Exascale: International Conference on Exascale Applications and Software, EASC 2014, Stockholm, Sweden, April 2-3, 2014, Revised Selected Papers, pp. 3 - 27 (Eds. Markidis, S.; Laure, E.). Springer, Cham (2015)
3. Kutzner, C.; Apostolov, R.; Hess, B.; Grubmüller, H.: Scaling of the GROMACS 4.6 molecular dynamics code on SuperMUC. In: Parallel Computing: Accelerating Computational Science and Engineering (CSE), pp. 722 - 730 (Eds. Bader, M.; Bode, A.; Bungartz, H. J.). IOS Press, Amsterdam (2014)
4. Sauguet, L.; Shahsavar, A.; Poitevin, F.; Huon, C.; Menny, A.; Nemecz, A. Haouz, A.; Changeux, J.-P.; Corringer, P.-J.; Delarue, M.: Crystal structures of a pentameric ligandgated ion channel provide a mechanism for activation. PNAS 111, 966-971 (2014)
5. Sauguet, L.; Poitevin, F.; Murail, S.; van Renterghem, C.; Moraga-Cid, G.; Malherbe, L.; Thompson, A. W.; Koehl, P.; Corringer, P.-J.; Baaden, M.; Delarue, M.: Structural basis for ion permeation mechanism in pentameric ligand-gated ion channels. EMBO J. 32, 728-741 (2013)
6. Kutzner, C.; Grubmüller, H.; de Groot, B. L.; Zachariae, U.: Computational Electrophysiology: The molecular dynamics of ion channel Permeation and selectivity in atomistic detail. Biophys. J. 101 (4), pp. 809 - 817 (2011)
Kutzner, C.; Ullmann, R.T.; de Groot, B.L.; Zachariaem U.; Grubmueller, H: Ions in Action - Studying Ion Channels by Computational Electrophysiology in GROMACS. https://doi.org/10.1016/j.bpj.2016.11.769
Prof. Dr. Helmut Grubmüller
Theoretical and Computational Biophysics Department
Max-Planck-Institute for Biophysical Chemistry
Am Fassberg 11, D-37077 Göttingen (Germany)