Molecular Dynamics Simulations of KirBac1.1 Mutants Reveal Global Gating Changes of Kir Channels

Abstract

Prokaryotic inwardly rectifying (KirBac) potassium channels are homologous to mammalian Kir channels. Their activity is controlled by dynamical conformational changes that regulate ion flow through a central pore. Understanding the dynamical rearrangements of Kir channels during gating requires high-resolution structure information from channels crystallized in different conformations and insight into the transition steps, which are difficult to access experimentally. In this study, we use MD simulations on wild type KirBac1.1 and an activatory mutant to investigate activation gating of KirBac channels. Full atomistic MD simulations revealed that introducing glutamate in position 143 causes significant widening at the helix bundle crossing gate, enabling water flux into the cavity. Further, global rearrangements including a twisting motion as well as local rearrangements at the subunit interface in the cytoplasmic domain were observed. These structural rearrangements are similar to recently reported KirBac3.1 crystal structures in closed and open conformation, suggesting that our simulations capture major conformational changes during KirBac1.1 opening. In addition, an important role of protein-lipid interactions during gating was observed. Slide-helix and C-linker interactions with lipids were strengthened during activation gating.

Figure 1

G143E_d location and induced channel opening. (A) Position of F146 (yellow) forming the helix bundle crossing gate and the introduced G143E_d mutant (orange). For clarity, only two opposing subunits are shown. (B) G143E is located in a tightly packed hydrophobic pocket formed by Y52 (slide-helix), F146 (yellow), L140, L144, and A147 (all four located in the TM2) of the same subunit (SU, colored gray) and V145, F146 (yellow), and F149 of the adjacent TM2 (colored light blue). (C) Averages of F146 Cα–Cα distances in WT and G143E_d simulations are shown as blue and green lines, respectively. Standard deviations are depicted as light shades accordingly. (D) Superposition of the TM2 helices of the open KirBac3.1 structure (pdb identifier: 3ZRS, shown in ocher) and the G143E_d mutant (final state, shown in green). The Cα atoms of F146 (KirBac1.1) and the equivalent Y132 (KirBac3.1) are shown as green and ocher spheres.

Figure 2

Conformational changes of F146 during gate opening. (A) χ1 angle distribution of the F146 side chain in WT (blue) and G143E_d (green) simulations. (B) Bottom view of the closed helix bundle crossing gate in WT simulations with F146 (yellow spheres) in the cavity facing conformation (χ1 angle of ∼160°). (C) Bottom view of an open helix bundle crossing gate in G143E_d simulations with F146 in the cavity lining rotameric state (χ1 angle of ∼270°). G143E_d are shown as orange spheres.

Figure 3

Water flux through the HBC gate. (A) Water impermeable gate in the WT simulation. Three SUs are shown for clarity. F103 residues are shown as yellow sticks. Water molecules are represented as spheres. (B) Water flux through the open gate in G143E_d simulations. G143E_d is depicted as orange sticks, and the K⁺ ion, as a purple sphere. (C) Water count of permeation events in the WT (blue line) and G143E_d (green shade) simulations.

Figure 4

Changes in the interaction network of the CTD. (A) Average of the CTD rotation angle in WT (blue) and G143E_d simulations (green). (B) Star graph of salt bridges between R271 and neighboring amino acids of adjacent SUs (aSU) and the same SU (sSU). Interactions in WT and G143E_d simulations are depicted as blue and green shades, respectively. The magnitude of interaction is normalized to the most prominent salt bridge in the protein. (C) Star graph of salt bridges between K191 and neighboring amino acids of aSUs and the sSU. (D) SU-interface of CTD conformation in WT simulation. The two aSUs are colored blue and gray. Salt bridges are depicted as dashed lines. (E) SU-interface of CTD conformation in G143E_d simulation. The two aSUs are colored green and gray. Salt bridges are depicted as dashed lines.

Figure 5

Salt bridge interactions of R153. (A) Star graph of salt bridges between R153 and neighboring amino acids of aSUs and the sSU. Interactions in WT and G143E_d simulations are depicted as blue and green shades, respectively. The magnitude of interaction is normalized to the most prominent salt bridge in the protein. (B) Starting conformation of G143E_d simulations. aSUs are colored green and gray. G143E_d and F146 are shown as orange and yellow sticks. (C) G143E_d–R153 salt bridge (dashed line) after 200 ns.

Figure 6

Average of the CTD rotation angle. Average rotation angle as a function of time in WT (blue), G143E_d (green), and G143E_d–R153A double mutant (magenta) simulations.

Figure 7

Analysis of G143E_p simulations. (A) Averages of F146 Cα–Cα distances in WT and G143E_p simulations are shown as blue and red lines, respectively. Standard deviations are depicted as light shades accordingly. (B) Average CTD rotation angle as a function of time in WT (blue) and G143E_p (red) simulations.

Figure 8

Free energy profile and corresponding gating changes of KirBac1.1 channel opening. (A) Energy profile along the main conformational changes of opening represented by the first eigenvector. Statistical error is depicted as green shading. (B) χ1 angle dynamics of F146 during gate opening. (C) Rotational angle of the CTD along the first eigenvector. (D) Occurrence of the salt bridge between R153 and G143E_d in all four SUs.

Figure 9

Protein–lipid interactions during gating. The total numbers of hydrogen bonds formed between lipids and R49 (A), K57 (B), and R151 (C) in all four WT and G143E_d simulations are depicted as blue and green lines, respectively.

Figure 10

Grow assay. The E. coli host strain was transformed with pQE60 vector, and the pQE60 vector carried KirBac1.1 WT, G143E, or G143C encoding DNA.