Asymmetric ligand binding facilitates conformational transitions in pentameric ligand-gated ion channels

Abstract

The anesthetic propofol inhibits the currents of the homopentameric ligand-gated ion channel GLIC, yet the crystal structure of GLIC with five propofol molecules bound symmetrically shows an open-channel conformation. To address this dilemma and determine if the symmetry of propofol binding sites affects the channel conformational transition, we performed a total of 1.5 μs of molecular dynamics simulations for different GLIC systems with propofol occupancies of 0, 1, 2, 3, and 5. GLIC without propofol binding or with five propofol molecules bound symmetrically, showed similar channel conformation and hydration status over multiple replicates of 100-ns simulations. In contrast, asymmetric binding to one, two or three equivalent sites in different subunits accelerated the channel dehydration, increased the conformational heterogeneity of the pore-lining TM2 helices, and shifted the lateral and radial tilting angles of TM2 toward a closed-channel conformation. The results differentiate two groups of systems based on the propofol binding symmetry. The difference between symmetric and asymmetric groups is correlated with the variance in the propofol-binding cavity adjacent to the hydrophobic gate and the force imposed by the bound propofol. Asymmetrically bound propofol produced greater variance in the cavity size that could further elevate the conformation heterogeneity. The force trajectory generated by propofol in each subunit over the course of a simulation exhibits an ellipsoidal shape, which has the larger component tangential to the pore. Asymmetric propofol binding creates an unbalanced force that expedites the channel conformation transitions. The findings from this study not only suggest that asymmetric binding underlies the propofol functional inhibition of GLIC, but also advocate for the role of symmetry breaking in facilitating channel conformational transitions.

FIGURE 1

Propofol binding sites in different GLIC systems. Top views of the transmembrane domains of (a) 0PFL; (b) 5PFL; (c) 3PFL; (d) 2PFL; (e) 1PFL; and (f) a side view of 5PFL. Propofol is in VDW representation and colored in purple.

FIGURE 2

Channel hydration under different scenarios of propofol binding. (a) Time evolution of the number of water molecules in the hydrophobic gate region (N_water). Three replicate runs, 5PFL-1, 5PFL-2 and 5PFL-3, are colored in green, red and black, respectively. Histograms of N_water were generated based on three replicate runs for each system, (b) 0PFL; (c) 5PFL; (d) 3PFL; (e) 2PFL; and (f) 1PFL. Snapshots with a 20-picosecond interval were taken from each run. A total of 15000 structures were used for each histogram analysis.

FIGURE 3

Distributions of lateral and radial tilting angles of TM2 for (a) 0PFL, (b) 5PFL, (c) 3PFL, (d) 2PFL, (e) 1PFL. (f) Depiction of radial and lateral directions for calculating the tilting angles. (g) The aligned crystal structures of the open-channel GLIC (PDB code: 3EAM; green) and the locally closed GLIC (PDB code: 3TLS; gray). The colors in (a) – (e) denote the channel hydration states associated with the TM2 tilting angles as defined by N_water: green for a fully hydrated channel (N_water >10); Purple for a partially dehydrated channel (0< N_water ≤10) and black for a fully dehydrated channel (N_water =0). Each system summarizes a total of 15,000 structures, sampled evenly over 100 ns for each of the three replicates. For comparison, a blue square and a blue triangle mark the TM2 tilting angles for the crystal structures of the open-channel GLIC and the locally closed GLIC, respectively. Counts of each hydration state for each system are provided in Fig. S5.

FIGURE 4

Top views of different conformations of the propofol-binding cavity. (a) A cavity (cyan surface) at the initial simulation; note that propofol (magenta sticks) is close to the edge of the cavity. (b) An expanded cavity after propofol penetrates deeper and residues T255 and Y197 are pushed apart over the course of the simulation. (c) A collapsed cavity (red surface) in the absence of propofol. When T255 and Y197 contact with each other, the large contiguous cavity is destroyed. The TM helices are labeled at the top of each helix.

FIGURE 5

Histograms of the propofol cavity volumes for (a) 0PFL; (b) 5PFL; (c) 3PFL; (d) 2PFL; and (e) 1PFL. Each histogram was generated based on the binding pockets in all five subunits over 3 replicated simulations. A total of 3000 structures and 31 bins were used in each histogram analysis to sample cavity volume from 0 to 300 ų for all of the systems. The maximum counts beyond the vertical scale limit are labeled in the plots.

FIGURE 6

Representative projections of the propofol-force trajectories, (a) 5PFL-1 and (b) 3PFL-1. The force trajectory (cyan) over a 100-ns simulation for each subunit is centered on propofol, which is marked by a black dot. The shape of the overall force trajectory is ellipsoid with the longest axis tangential to the pore. The first (red arrows) and second (purple arrows) principal components of each force trajectory are scaled by their respective eigenvalues and projected onto the same plane as the force trajectory. Zoom-in views of the propofol force on individual residues Y197 and T255 of subunit B in (a) 5PFL and (b) 3PFL were generated based on the force calculation separately for each residue. The force trajectory is colored in blue and red for the first and last 50-ns simulation, respectively. The coordinate trajectory of the propofol’s center of mass is shown in green and black for the first and last 50-ns simulation, respectively. The time averaged net force on Y197 and T255 for first 50-ns and entire 100 ns simulations are shown in orange and yellow arrows, respectively. Reference scales for the amplitude of the force in the overall and zoom-in views are shown.