Atomistic simulation of voltage activation of a truncated BK channel

Abstract

Voltage-dependence gating of ion channels underlies numerous physiological and pathophysiological processes, and disruption of normal voltage gating is the cause of many channelopathies. Here, long timescale atomistic simulations were performed to directly probe voltage-induced gating transitions of the big potassium (BK) channels, where the voltage sensor domain (VSD) movement has been suggested to be distinct from that of canonical Kv channels but remains poorly understood. Using a Core-MT construct without the gating ring, multiple voltage activation transitions were observed at 750 mV, allowing detailed analysis of the activated state of BK VSD and key mechanistic features. Even though the S4 helix remains the principal voltage sensor in BK, its vertical displacement is only ~3 Å and accompanied by significant lateral movements. The nature of the predicted VSD movement is in strong agreement with recent Cryo-EM structural studies of mutant BK channels with constitutively activated VSD. Free energy analysis based on the predicted activation transition yielded a total gating charge of 0.44 e per VSD, consistent with the experimental range of 0.48-0.65 e. We further show that the ability of modest physical movements with a small total gating charge to drive effective voltage gating of BK can be attributed to large gradients in the local electric field as reshaped by the protein. Furthermore, the S4 movement is coupled to the pore opening through a non-canonical pathway that involves the tightly packed S4-S5-S6 interface. These distinct mechanistic features may be relevant to voltage gating of other ion channels where VSDs are not domain-swapped with respect to the pore-gate domain.

Keywords: computational biology; conformational transition; electric field reshaping; free energy; gating charge; hydrophobic gating; molecular biophysics; none; pore-sensor coupling; structural biology; systems biology.

Figure 1.. Overall structures of BK and Kv channels.

(a) The overall structure of BK channels in the Ca²⁺-free state (PDB: 6v3g) with key domains and regions labeled. Each monomer is shown in the same color. (b) Top view of the TMD of BK channels, showing the non-domain-swapped VSD/PGD configuration. (c) Top view of the TMD of Kv 1.2 channel (PDB: 3lut), where the VSDs and PGDs are domain-swapped. Note the much tighter packing of S4 from VSD and S5/S6 from PGD in BK channels.

Figure 2.. Voltage activation of Core-MT BK channels.

(a–d) Results from a 10-μs simulation under 750 mV (sim2b in Supplementary file 1). Each data point represents the average of four subunits for a given snapshot (thin gray lines), and the colored thick lines plot the running average. (a) z-displacement of key side chain charged groups from initial positions, (b) z-displacement of centers-of-mass of VSD helices from initial positions, (c) backbone RMSD of the pore-lining S6 (F307-L325) to the open state, and (d) tilt angles of all TM helices. The locations of charged groups were taken as those of guanidinium CZ atoms (for Arg) and side chain carboxyl carbons (for Asp/Glu). Only residues 313–324 of S6 were included in tilt angle calculation, and the values in the open and closed Cryo-EM structures are marked using purple dashed lines for reference in panel d. (e) Superimposition of the initial (0 μs) and final (10 μs) structures of the pore (residues F307-L325; red cartoon), in comparison to the open Cryo-EM structure (cyan cartoon). The view shown is from the bottom (cytosolic side). (f) Average pore profiles calculated from the first and last 0.1 μs of sim2b, with error bars showing standard error. The pore profile derived from PDB 5tj6 (open state) is shown as a reference.

Figure 2—figure supplement 1.. The charged group z-displacement, z-displacement of the centers of mass of TM helices, number of pore waters, the averaged pore profiles during four Anton 2 simulations of Core-MT BK channels at 0 mV (row 1), 750 mV (rows 2–3), and 300 mV (rows 4–6) membrane voltages.

See Materials and methods for additional details of the simulation and analysis.

Figure 2—figure supplement 2.. Distribution of lipid phosphor atoms under different membrane voltages.

The distributions were derived from the last 500 ns of simulation in 0 mV (sim 1), 300 mV (sim 9), and 759 mV (sim 2 a). Only the phosphate groups not within 12 Å of any protein atoms were selected for analysis. The average distances between phosphor atoms in the upper and lower leaflets are 45.6 Å, 47.0 Å, and 47.0 Å at 0 mV, 300 mV, and 750 mV, respectively. One-way ANOVA suggests that the difference in average membrane thickness under different voltages is not significant.

Figure 3.. Pore rehydration and ion conductance.

(a) The number of water molecules inside the pore as a function of time during simulation (sim 2b), with the upper panel showing recorded ion permeation events. Inserts show snapshots of the pore region at three representative time points. (b) Snapshots illustrating key steps of K⁺ ions passing through the filter. Potassium ions inside or near the filter are colored according to their identities. The water molecule bridging two ions inside the filter is also shown as van der Waals spheres.

Figure 3—figure supplement 1.. Conductance of the open state of Core-MT derived from the Ca²⁺-bound full-length BK structure (PDB: 5tj6).

The voltage was set at 750 mV and ion permeation events are shown as impulses. The conductance estimated from the second half of the trajectory is ~6 pS.

Figure 4.. VSD gating charge and voltage-sensing movements.

(a) Average voltage-induced movements of key charges along the membrane normal (z-axis) with respect to the initial resting state structure, derived from last 500 ns of 750 mV simulation sim 2b. Error bars show the standard deviations. (b) Conformations of key charged residues in the resting (silver) and activated (orange) states of BK VSD. The resting and activated states are represented using the snapshots at 0 and 10 μs of sim2b, respectively. (c) Distributions of centers of mass of TM helices along the membrane lateral directions (x and y) (view from the cytosolic side). The distributions for resting and activated states were derived from the first and last 500 ns of the 750 mV simulation sim 2b, respectively, which were converted into the free energy scale by ~R T ln P(x,y) with T=300 K. The contour for the resting state distribution (dotted lines) is drawn at 4 kcal/mol. (d) Overlay of the resting (silver) and activated (orange) states of the TM domain of the Core-MT BK channel. The green and red spheres mark the backbone Cα atom of S4 R213 in the resting and activated states. Only one subunit is shown for clarity, but all four filter loops are shown for reference.

Figure 4—figure supplement 1.. Steered MD simulations of BK activation.

(a) Illustration of the steered MD setup. Initially, the reference point (orange circle) is placed at the position of R210 or R213 guanidinium CZ atom in the resting state (black circle/blue rectangle). From 0 to 100 ns, the reference point moves along the z-axis (orange arrow) from the resting state position towards the activated state position (red rectangle). A harmonic positional restraint of 5 kcal/mol·Å², applied along the z-axis only, was used between the reference point and the CZ atom, steering the CZ atom to move along the z-axis with the reference. After 100 ns, the reference point remains fixed at the activated state position (red rectangle), allowing the VSD and the rest of the channel to respond to R210 and R213 movements. (b and c) Numbers of pore water and S6 tilting angle as a function of simulation time during two of the steer MD simulations that lead to pore opening (black trace: replica 2; red trace: replica 4). (d) Overlay of the pore structure at the end of replica 2 (red) and the open Cryo-EM structure (cyan). (e) Pore profile at the end of replica 2 (red) in comparison to those from the initial state (dashed line) and the open Cryo-EM structure (black).

Figure 5.. Electrostatic potential fields of the Core-MT BK channel at 750 mV with resting and activated VSDs.

The fields were calculated as the averages of the first (resting) and last (activated) 250 ns of simulation sim2b. The fields are shown on a plane that goes through the filter and R210. Only two subunits of Core-MT BK are shown for clarity, with side chains of key S4 charges shown in sticks.

Figure 6.. Dynamic community, coupling pathways, and information flow of VSD-pore coupling in BK.

(a) Dynamic community analysis showing that TM S4-6 are clustered into single tightly coupled community (blue network). The nodes (residues) and edges (contacts) are colored based on the community number. (b) Optimal and suboptimal pathways of dynamic coupling between R213 (VSD S4) and E321 (pore-lining S6). All paths are colored green except for the optimal path, which is colored red. Nodes with information flow value >0.02 are highlighted in purple; (c) Information flow profile of the Core-MT BK channel with R213 as the source and E321 as the sink node (labeled by red circle), respectively. All dynamic coupling analysis was derived from the last 500 ns of sim 1 (closed state at 0 mV; see Supplementary file 1).

Figure 6—figure supplement 1.. Covariance matrices of the Core-MT BK channel before (left) and after (right) voltage-induced activation (sim 2b).

Figure 6—figure supplement 2.. Locking and concerted movements of S4, S5, and S6.

(a) The difference of residue-residue contact probability between the closed (sim 1, 0–0.5 μs) and activated state (sim 2b, 9.5–10 μs). (b) Side and (c) bottom view of the average activated (cyan) and closed (white) structures. The structures are aligned using the filter and P-loop. Dashed lines depict contacts where the probability increased (blue) or decreased (red) by more than 0.4 after activation. The Cα atoms of R207, R210, and R213 are represented as spheres. Residues (K234, L235, E321, and E324) involved in S6 bending are illustrated as yellow and red sticks for closed and activated states, respectively; (d) Contact maps of S4-6 before (left) and after (right) voltage activation of Core-MT BK channels.

Figure 6—figure supplement 3.. Dynamic community, coupling pathways, and information flow of VSD-pore coupling derived from simulation of Ca²⁺-bound structure.

All results were derived from the last 200 ns of sim 7 (Supplementary file 1). (a) Dynamic community analysis showing that TM S4-6 are clustered into a single tightly coupled community (blue network). The nodes (residues) and edges (contacts) are colored based on the community number. (b) Optimal and suboptimal pathways of dynamic coupling between R213 (VSD S4) and E321 (pore-lining S6). All paths are colored green except for the optimal path, which is colored red. (c) Information flow profile with R213 as the source and E321 as the sink node (labeled by red circle), respectively. Dynamic coupling analysis from the last 500 ns of sim 1 (closed state at 0 mV) is also shown as reference (black trace).

Figure 6—figure supplement 4.. Changes in residue-residue contact probability relative to the closed channel.

(a) S4-S5 interface, and (b) S5-S6 interface. The contact probabilities of the closed state were derived from sim 1 (0–0.5 μs). Black bars show the changes observed in the voltage-driven activated state from sim 2b (9.5–10 μs), and red bars show the changes in the cryo-EM open state from sim 7 (200–400 ns). Only residue pairs with a contact probability change of at least 0.4 in either sim 2b or sim 7 are shown for clarity.