PIP₂-dependent coupling is prominent in Kv7.1 due to weakened interactions between S4-S5 and S6

Abstract

Among critical aspects of voltage-gated potassium (Kv) channels' functioning is the effective communication between their two composing domains, the voltage sensor (VSD) and the pore. This communication, called coupling, might be transmitted directly through interactions between these domains and, as recently proposed, indirectly through interactions with phosphatidylinositol-4,5-bisphosphate (PIP₂), a minor lipid of the inner plasma membrane leaflet. Here, we show how the two components of coupling, mediated by protein-protein or protein-lipid interactions, both contribute in the Kv7.1 functioning. On the one hand, using molecular dynamics simulations, we identified a Kv7.1 PIP₂ binding site that involves residues playing a key role in PIP₂-dependent coupling. On the other hand, combined theoretical and experimental approaches have shown that the direct interaction between the segments of the VSD (S4-S5) and the pore (S6) is weakened by electrostatic repulsion. Finally, we conclude that due to weakened protein-protein interactions, the PIP2-dependent coupling is especially prominent in Kv7.1.

Figure 1. Identification of a PIP₂ putative binding site in Kv7.1.

(a) The average number of salt bridges formed between each PIP₂ molecule (in total 38) and positive residues of the channel exposed to the interface between the lower bilayer leaflet and the solution. Red bars correspond to PIP₂ molecules bound at the intrasubunit binding site. (b), (c) The PIP₂ intrasubunit binding site is located at the VSD/PD interface of Kv7.1. The channel in its activated/open state is present in ribbons. Color code: the VSD and the PD are shown in white and green respectively; S4 is highlighted in cyan. PIP₂ molecules are present as sticks with white hydrocarbon tails. Oxygen and phosphorus atoms of PIP₂ headgroups are colored in red and tan. For clarity, only the phosphorus (tan) and nitrogen (blue) atoms of the POPC bilayer headgroups are shown. The snapshot was taken after 50 ns of the equilibration. (b) and (c) correspond to the bottom and side views respectively.

Figure 2. There are two modes of protein-lipid interactions each favored either in the activated/open (A/O, top panels) or resting/closed (R/C, bottom panels) states of Kv7.1.

(a) Probabilities of salt bridges formation between PIP₂ and positive residues of Kv7.1 among all subunits. Multiple MD runs are considered. The standards deviations (SDs) are represented as bars. (b) Coordination of PIP₂ by residues composing the intrasubunit binding site: K183, R190 and R195 of the S2–S3 loop, R237 and R243 of S4, R249 of the S4–S5 linker, K354 and K358 of the S6 terminus. K183 and R249, the gain-of-function residues, are highlighted in red. (c) Side (left panel) and top (right panel) views of the representative equilibrium position of PIP₂ at the intrasubunit binding site. Note that PIP₂ anchors the S6 terminus in the A/O state of Kv7.1. When the channel is R/C, the lipid is shifted toward the VSD to interact with the bottom of S4.

Figure 3. Mutations of the Kv7.1 gain-of-function residues, K183 and R249, shift the position of PIP₂ toward S6 in both the A/O (left panel) and R/C (right panel) states of the channel.

(a) Probabilities of salt bridges formation between PIP₂ and residues of Kv7.1. (b) Shift of PIP₂ (grey arrow) between positions in the WT (blue spheres, representing the P1, P4 and P5 atoms of the four PIP₂ molecules) and in the gain-of-function mutants (red spheres). Here, the systems with the most significant shifts are reported: R249E in the A/O state (left panel: side and top views) and K183E in the R/C state (right panel: side and top views).

Figure 4. Several positive residues of S4–S5 and S6 modulate the strength of the S4–S5/S6 interactions.

(a) Per-residue decomposition analysis of the nonbonded energy of interaction between S4–S5 and S6 performed in the A/O (top) and R/C (bottom) states of the WT channel. The residues are colored corresponding to their contribution. R249 and K358 shown in blue provide a positive (repulsive) contribution. Several residues highlighted in red stabilize the interaction between these two regions. (b) The nonbonded energies of the S4–S5/S6 interactions of R249Q/E (top panel), K358E (middle panel) and K183Q/E (bottom panel) compared to the WT's one reported for the A/O (left) and R/C (right) states of Kv7.1. Here vdW, El and Tot correspond to the van-der-Waals, electrostatic and total nonbonded energy respectively. The error bars (SE) are present.

Figure 5. The protein-protein and protein-lipid components of coupling detected in K358E.

(a) Normalized F/V curves of K358E (red solid) and K358E/L353K (red dashed) expressed alone (left) or with CiVSP (right). Normalized F/V curves of the WT (black solid) and L353K (black dashed) with and without CiVSP are shown for comparison. (b) Normalized F/V curves of K358E expressed alone or with CiVSP. (c) Normalized F/V (solid) and G/V (dashed) curves of K358E (red) and the WT (black). Note that the latter are superimposed. The WT data reproduced from.

Figure 6. Mutagenesis of basic residues of the S4–S5 linker and the S6 terminus indicated the importance of electrostatic interactions between these two regions.

(a) Normalized currents of the WT (black), charge neutralizing (green – R249Q, K358N) and charge reversing (red – R249E, K358E) mutations (top panel). The G/V curves of the WT (black), R249Q (green solid), R249E (red solid, ΔΔG = 0.56 ± 0.14 kcal/mol), K358N (green dashed), K358E (red dashed, ΔΔG = 0.68 ± 0.07 kcal/mol) and R249E/K358E (blue solid, ΔΔG = 2.60 ± 0.13 kcal/mol) (bottom panel). (b) Normalized currents (top panel) and the G/V curves (bottom panel) of the WT (black), K354E (blue), K358E (red), R360E (cyan), K362E (magenta) and R366E (orange). (c) Normalized currents (top panel) and the G/V curves (bottom panel) of the WT (black), K183N (green).