An energy-efficient gating mechanism in the acetylcholine receptor channel suggested by molecular and Brownian dynamics

Abstract

Acetylcholine receptors mediate electrical signaling between nerve and muscle by opening and closing a transmembrane ion conductive pore. Molecular and Brownian dynamics simulations are used to shed light on the location and mechanism of the channel gate. Four separate 5 ns molecular dynamics simulations are carried out on the imaged structure of the channel, a hypothetical open structure with a slightly wider pore and a mutant structure in which a central ring of hydrophobic residues is replaced by polar groups. Water is found to partially evacuate the pore during molecular simulations of the imaged structure, whereas ions face a large energy barrier and do not conduct through the channel in Brownian dynamics simulations. The pore appears to be in a closed configuration despite containing an unobstructed pathway across the membrane as a series of hydrophobic residues in the center of the channel provide an unfavorable home to water and ions. When the channel is widened slightly, water floods into the channel and ions conduct at a rate comparable to the currents measured experimentally in open channels. The pore remains permeable to ions provided the extracellular end of the pore-lining helix is restrained near the putative open configuration to mimic the presence of the ligand binding domain. Replacing some of the hydrophobic residues with polar ones decreases the barrier for ion permeation but does not result in significant currents. The channel is posited to utilize an energy efficient gating mechanism in which only minor conformational changes of the hydrophobic region of the pore are required to create macroscopic changes in conductance.

FIGURE 1

Channel structure and simulation system. (A) The system used for MD simulations is shown with the front half of the atoms removed to make the pore visible. The transmembrane protein (yellow), shown in surface representation, spans the POPC lipid membrane (brown) and is surrounded by water (blue) as well as Na⁺ (purple) and Cl⁻ (green) ions. The entire system contains 66,462 atoms. (B) The transmembrane portion of the protein is comprised of five subunits each containing four membrane-spanning α-helices. The imaged structure viewed looking through the pore is shown in blue, and the putative open state structure is indicated in red. (C) The radius of the imaged (blue) and putative open (red) pore plotted for the transmembrane and ligand binding protein domains. The approximate axial location of the membrane is indicated by the orange bar.

FIGURE 2

Configuration of water in the transmembrane pore during MD simulations. The positions are shown at the beginning and 2.5 ns into simulations of the imaged structure (A and B); putative open structure (C and D); and V255T mutant channel (E and F). The atomic surface of the protein is shown with the front half of the atoms removed with hydrophobic regions shown in white, polar regions in green, and acidic and basic residues shown in red and blue, respectively. To indicate the edge of the pore, the region where the protein meets the clipping plane has been highlighted in yellow. The exact shape of the pore and protein edge are highly dependent on the orientation of the clipping plane.

FIGURE 3

RMSD of backbone carbon atoms during MD simulations. Calculations are made for the entire protein (A) and M2 helix (B) during the simulations of the imaged (blue), putative open (red), pinned open (black), and V255T mutant (green) structures respectively.

FIGURE 4

Water density and likely channel state. The density of water in the central 5 Å portion of the pore normalized by the bulk water density is plotted during MD simulations (blue) for the imaged pore structure (A); putative open structure (B); pinned open structure (C); and V255T mutant channel (D). The likely channel state is also shown (red, right-hand scale), assuming that the channel is closed whenever the water density in the pore falls below 0.65 times the bulk value.

FIGURE 5

Pore radius. The radius of the pore as found using the program HOLE is plotted for the transmembrane portion of the protein at the beginning (solid line), middle (dotted line), and end (dashed line) of 5 ns MD simulations. Radii are plotted for the imaged structure (A), putative open structure (B), pinned open structure (C), and V255T mutant structure (D). The radius of the hypothetical open structure is also illustrated by the shaded lines in A and D for comparison, and the radii of the initial imaged structure is shown by the shaded lines in B and C.

FIGURE 6

Energy barriers to ion conduction. The electrostatic potential energy is plotted for a Na⁺ ion moving through the transmembrane pore, and 10 additional ions are allowed to find their minimum energy positions in the extracellular end of the pore and ligand binding domain. (A) The barrier is calculated for the imaged structure at the start of simulations (solid line); the structure after 2.5 ns of MD simulation (dashed line); and the V255T mutant structure after 2.5 ns of simulation (dashed-dotted line). Also shown is the energy barrier presented to a Cl⁻ ion using the imaged structure of the channel after 2.5 ns of simulation (dashed-dotted-dotted line). (B) The energy barrier to Na⁺ conduction in the putative open channel at the beginning (dashed-dotted line) and after 2.5 ns MD simulations (dashed line); and the pinned open structure after 2.5 ns of simulation (solid line) are illustrated.

FIGURE 7

Ion dwell histogram. The locations where ions dwell in the transmembrane channel and extracellular binding domain during BD simulations are illustrated. Results are shown for Na⁺ ions (black bars) and Cl⁻ ions (white bars) in the imaged structure (A), hypothetical open structure (B), and pinned open structure (C).

FIGURE 8

Conformations of the M2 helices. The conformation of the M2 helices as well as the side chains of the central hydrophobic residue (VAL255 in the α-subunit and equivalent residues on the other subunits) are illustrated at the beginning (A) and end (B) of the simulation of the imaged structure; the beginning (C) and end (D) of the simulation of the putative open structure; the end of the simulation of the pinned open structure (E) and the end of the simulation of the V255T mutant structure (F). The configuration of the M2 helix at the start of each simulation is indicated by the gray tubes.

FIGURE 9

Total energy of the system during MD simulations. The total energy is plotted for the simulation of the imaged (blue), putative open (red), and pinned open (black) structures, sampling the energy every pS. Average energy values, in which the data is averaged in a moving 25 pS window, are shown.