Initial response of the potassium channel voltage sensor to a transmembrane potential

Abstract

Early transition events of the voltage sensor (VS) of Kv1.2 potassium channel embedded in a lipid membrane are triggered using full atomistic molecular dynamics (MD) simulations. When subject to an applied hyperpolarized transmembrane (TM) voltage, the VS undergoes conformational changes and reaches a stable kinetic intermediate state, beta', within 20 ns. The gating charge ( approximately 2e) associated with this fast transition results mainly from salt-bridge rearrangements involving negative charges in S2 and S3 and all but the two top residues R(294) and R(297) of S4. Interactions of the latter with phosphomoieties of the lipid head groups appear to stabilize the kinetic state beta'.

Figure 1

Left: initial MD configuration of Kv1.2 embedded in a POPC bilayer (grey) and ~150 mM KCl solution (VS: green, S4: yellow, pore: red, T1 intracellular tetramerization domain: blue, K⁺: orange, Cl⁻: cyan). For clarity, 2 of 4 channel subunits are depicted and water is not shown. The MD simulation box is drawn in light blue. Note that the electrolytes extend up to the air/water interfaces. A net charge imbalance between the upper and lower electrolytes created by displacing cations from the lower to the upper bath induces a finite TM potential. Right: One of the four voltage-sensor domains highlighting the charged residues (blue) of the S4 helix: from top to bottom: R²⁹⁴ R²⁹⁷ R³⁰⁰ R³⁰³ K³⁰⁶ and R³⁰⁹.

Figure 2

Top: Electrostatic potential profile Φ(z) along the bilayer normal (z) of the initial configuration. Φ(z) is derived directly from the MD simulation as a double integral of the charge distribution of all atoms averaged ,over the membrane plans, ρ(z), as $\mathrm{\Phi }(z)-\mathrm{\Phi }(0)=-{\epsilon }_0^{-1}{\displaystyle \int {\displaystyle \int \rho (z")dz"dz\text{'}.}}$ Here, as a reference Φ is set to zero in the upper electrolyte. The arrows indicate the approximate position of the lipid-water interfaces. Bottom: Relaxation of the TM potential during MD run. Every point corresponds to the TM voltage estimated from the average over a simulation-time window of ~ 1.2 ns - the error associated with each estimate is ± 50 mV. The ΔV reduction provides a direct estimates of the gating charge (see main text).

Figure 3

Initial (left) and final (right) conformations of a VS domain that are representative of respectively the α and β’ states. The arrow highlights the collective motion of the lower charged residues of S4. Note the specific salt bridges between basic residues (blue) of S4 (yellow) and acidic residues (red) of segments S1, S2 and S3 (green).

Figure 4

Top: Electrical distances ${\delta }_i^{\lambda }$ for each TM residue in the α (black) and β′ (orange) conformations and net charge per residue (green) along the Kv channel sequence (excluding the T1 domain). The position of the TM segments S1 to S6 and the P loop are indicated by arrows. ${\delta }_i^{\lambda }$ were normalized assuming ${\delta }_i^{\lambda }=1$ and 0 for residues positioned respectively above 25 Åand below −25 Å from the bilayer centre. Bottom: Cumulative (yellow line) and per-residue (bars) gating charges. The S4-basic residues (blue) and the VS-negative residues (red: E157, E183, E226, E236 and D259) are highlighted. Contributions of the mobile loop residues (gray) were not included in the estimates of the cumulative charge.