Pore- and voltage sensor-targeted KCNQ openers have distinct state-dependent actions

Abstract

Ion channels encoded by KCNQ2-5 generate a prominent K⁺ conductance in the central nervous system, referred to as the M current, which is controlled by membrane voltage and PIP2. The KCNQ2-5 voltage-gated potassium channels are targeted by a variety of activating compounds that cause negative shifts in the voltage dependence of activation. The underlying pharmacology of these effects is of growing interest because of possible clinical applications. Recent studies have revealed multiple binding sites and mechanisms of action of KCNQ activators. For example, retigabine targets the pore domain, but several compounds have been shown to influence the voltage-sensing domain. An important unexplored feature of these compounds is the influence of channel gating on drug binding or effects. In the present study, we compare the state-dependent actions of retigabine and ICA-069673 (ICA73, a voltage sensor-targeted activator). We assess drug binding to preopen states by applying drugs to homomeric KCNQ2 channels at different holding voltages, demonstrating little or no association of ICA73 with resting states. Using rapid solution switching, we also demonstrate that the rate of onset of ICA73 correlates with the voltage dependence of channel activation. Retigabine actions differ significantly, with prominent drug effects seen at very negative holding voltages and distinct voltage dependences of drug binding versus channel activation. Using similar approaches, we investigate the mechanistic basis for attenuation of ICA73 actions by the voltage-sensing domain mutation KCNQ2[A181P]. Our findings demonstrate different state-dependent actions of pore- versus voltage sensor-targeted KCNQ channel activators, which highlight that subtypes of this drug class operate with distinct mechanisms.

Figure 1.

The ICA73 is excluded from resting KCNQ2 channels. (A) Conductance–voltage relationships for WT KCNQ2 channels expressed in Xenopus laevis oocytes in various concentrations of ICA73, normalized to peak current in control for each cell. The KCNQ2 gating parameters were as follows: control V_1/2 = −48.7 ± 1.0 mV, k = 9.9 ± 0.4 mV; 10 µM ICA73 V_1/2 = −76.4 ± 2.2 mV, k = 17.9 ± 0.5 mV; and 30 µM ICA73 V_1/2 = −108.6 ± 3.0 mV, k = 21.3 ± 0.7 mV, n = 10. (B) Exemplar patch clamp records of WT KCNQ2 channels expressed in HEK293 cells. In control solution, cells were held at −120 mV, with depolarizations to +20 mV. 10 µM ICA73 was applied extracellularly for ∼1 min while holding at −120 mV. Cells were then depolarized again to +20 mV (first pulse in ICA73), with repeated depolarizations every 6 s (subsequent pulses in ICA73). (C) Instantaneous current (marked by asterisks in B) in the first pulse and subsequent pulses as a percentage of the peak current, calculated from cells which had drug applied during holding potentials of −120, −80, and −60 mV (n = 4–7). Hollow symbols are data from individual cells, and filled symbols are mean ± SEM. (D) Exemplar patch clamp records of KCNQ2 channels expressed in HEK293 cells, incubated in 10 µM ICA73 for 1 min. Cells were held at −100 mV before being depolarized to +20 mV for 200 ms (first sweep) and then hyperpolarized to −120 mV for 10 s. The duration of the depolarizing pulse increased by 500 ms in each subsequent sweep. (E) Tail currents at −120 mV (from D) were normalized to better visualize kinetic differences.

Figure 2.

The ICA73 binding kinetics correlate with KCNQ2 channel activation. (A) Voltage protocol and exemplar current traces for KCNQ2 channels in HEK293 cells. Cells were held at −80 mV and pulsed to a range of voltages between −120 and 40 mV for 5 s. After 1 s at the test voltage, 100 µM ICA73 was applied rapidly. (B) Magnified view of the current traces immediately following ICA73 application, normalized to the peak current at −20 mV. Single exponential fits are superimposed in blue. Occasionally, a sum of two exponential equations was required (red fit), in which case the dominant component was used as a measure of interaction rate. Fits also excluded the first 30 ms after the solution exchange, which is contaminated by rapid block by DMSO. (C) Exemplar traces for WT KCNQ2 channels in HEK293 cells, with DMSO rapidly applied. Cells were held at −100 mV, then pulsed to a range of voltages between 0 mV and −100 mV for 2.5 s before DMSO application. Cells were then pulsed to −60 mV for 5 s. (D) Rates of ICA73 binding (green circles, left axis; n = 4–6) and KCNQ2 conductance-–voltage relationship in HEK293 cells (dotted line, right axis). Error bars represent mean ± SEM.

Figure 3.

The ICA73 accessibility shifts to negative voltages in a shifted KCNQ2 channel mutant. (A) Conductance–voltage relationships for KCNQ2 [R201A] channels expressed in HEK293 cells in control (black circles) or 100 µM ICA73 (green circles). The voltage dependence of WT KCNQ2 channel opening in control is indicated by the dashed line. (B) Exemplar traces for WT KCNQ2 (black) and R201A (green) after rapid application of 100 µM ICA73 at −120 mV. The rates of ICA73-mediated current potentiation at −60 (WT), −120, and −140 mV (R201A) are also shown for comparison. (C) Rates of ICA73 binding to WT KCNQ2 (filled circles; n = 4–6) and R201A (unfilled circles; n = 3–6) are plotted against membrane voltage. Error bars in A and C represent mean ± SEM.

Figure 4.

Retigabine can access open and closed channels. (A) Conductance–voltage relationships for WT KCNQ2 channels expressed in X. laevis oocytes in various concentrations of retigabine (RTG), normalized to peak current in control for each cell. The KCNQ2 parameters of activation were as follows: control V_1/2 = −36.8 ± 1.1 mV, k = 9.2 ± 0.4 mV; 10 µM RTG V_1/2 = −47.6 ± 2.0 mV, k = 12.2 ± 1.1 mV; 30 µM RTG V_1/2 = −69.5 ± 5.4 mV, k = 21.6 ± 2.0 mV; and 100 µM RTG V_1/2 = −81.7 ± 3.6 mV, k = 20.7 ± 1.4 mV, n = 6. Error bars represents mean ± SEM. (B) Wash-in experiment described in Fig. 1 B was performed using 30 µM RTG. Records depict the experiment with a holding potential of −120 mV. (C) Instantaneous current as a percentage of peak current, calculated from cells which had drug applied at holding potentials of −120, −80, and −60 mV (n = 4–6). Hollow symbols are data from individual cells, and filled symbols represent the mean ± SEM. (D and E) Exemplar patch clamp records of WT KCNQ2 channels expressed in HEK293 cells, in control (D) or after incubation in 30 µM RTG for 1 min (E). Cells were held at a holding potential of −100 mV before being depolarized to +20 mV for 200 ms (first sweep), then hyperpolarized to −120 mV for 5 s. The duration of the depolarizing pulse increased by 500 ms in each subsequent sweep. (F) Current traces from D and E were normalized to better visualize kinetic differences.

Figure 5.

Retigabine interacts with resting channels. (A) The voltage and timed perfusion protocol described in Fig. 2 A was performed using 100 µM RTG. Exemplar current traces from cells expressing WT KCNQ2 channels are shown. (B) Magnified view of the current traces immediately following RTG addition, normalized to peak current. Single exponential fits are superimposed in blue. Occasionally, a sum of two exponential equations was required (green fits), in which case the dominant component was used as a measure of interaction rate. Fits also excluded the first 30 ms after the solution exchange, which is contaminated by rapid block by DMSO as described in Fig. 2 C. (C) Rates of RTG binding (red circles, left axis; n = 2–7) and KCNQ2 conductance–voltage relationship in HEK293 cells (dotted line, right axis). Error bars represent mean ± SEM.

Figure 6.

Acceleration of KCNQ2 activation after retigabine exposure at negative voltages. (A) 5 µM retigabine was applied to HEK 293 cells expressing WT KCNQ2 channels, for varying durations (0, 1, or 2 s) at −100 mV, followed by a voltage step to −40 mV. (B) Identical solution and voltage steps were applied as in A but with 5 µM ICA73. (C) Summary data illustrating the 50% rise time of current upon depolarization to −40 mV for ICA73 (n = 3) or retigabine (n = 5), illustrating retigabine-mediated acceleration of channel activation after a 2-s exposure at −100 mV (Student’s t test compared the 0- and 2-s exposures in each drug; *, P < 0.05 relative to the 0-s exposure). Error bars represent mean ± SEM.

Figure 7.

Differential effects of RTG and ICA73 on KCNQ2 activation kinetics. (A and C) Exemplar patch clamp records of WT KCNQ2 channels before and after exposure to RTG or ICA73. Under control conditions, cells were held at −100 mV, depolarized to 0 mV, and repolarized to −100 mV while drug was applied (1-min wash-in). Cells were then depolarized to 0 mV (2 s) and returned to −100 mV for varying durations (Δtime) to allow partial channel closure. (B and D) Exemplar current traces showing kinetics of the activating fraction of current in RTG or ICA73 (red and green, respectively, after varying durations of repolarization) or control solution (black), normalized to the peak current after drug. Activation time constants in control or drug conditions (n = 7) are depicted on the right. Hollow symbols are data from individual cells, and filled symbols are mean ± SEM.

Figure 8.

The ICA73 binding to KCNQ2[A181P] correlates with channel activation. (A) The voltage and timed perfusion protocol described in Fig. 2 A was performed using 100 µM ICA73 with KCNQ2[A181P] channels. Exemplar current traces from cells expressing KCNQ2[A181P] channels are shown. (B) Magnified view of current traces immediately following drug addition, normalized to highlight interaction kinetics with ICA73. (C) Rates of ICA73 binding (green circles, left axis; n = 2–6) and KCNQ2[A181P] conductance–voltage relationship in HEK293 cells (dotted line, right axis). Error bars represent mean ± SEM.

Figure 9.

The ICA73 binding is not affected in KCNQ2[A181P] channels. (A) Model of KCNQ2 channel structure, highlighting the location of A181 in the voltage-sensing domain (highlighted in blue), distant from W236 that forms the canonical RTG binding site in the pore domain (yellow). (B) Exemplar current traces for WT KCNQ2 channels (black) and KCNQ2[A181P] channels (blue). Cells were held at −100 mV and depolarized to 0 mV, during which time 30 µM ICA73 was rapidly applied extracellularly. (C) Activation time constants of channel opening by a voltage step to 0 mV and following rapid drug application at 0 mV (n = 5–6). Hollow symbols are data from individual cells, and filled symbols are mean ± SEM.

Figure 10.

Accelerated ICA73 unbinding from KCNQ2 [A181P] channels. (A and B) Exemplar current recordings for WT KCNQ2 (black) and KCNQ2[A181P] (blue), following rapid washout of 30 µM ICA73 from bath at 0 (A) or −100 mV (B). (C) Time constants of current decay at 0, −50, −100, and −150 mV after rapid ICA73 washout (n = 4–6). Also depicted are deactivation tail current kinetics between −90 and −130 mV after a depolarizing prepulse to 0 mV (tail current kinetics; n = 5). Hollow symbols are data from individual cells, and filled symbols are mean ± SEM.

Figure 11.

Distinct binding sites and state dependence of RTG and ICA73. (A) Schematic diagram of the mechanism of action of ICA73, targeting a binding site in the voltage-sensing domain. The ICA73 is proposed to interact nearly exclusively with the activated conformation of the voltage sensor, leading to stabilization of the activated state. (B) Schematic diagram of the mechanism of action of RTG, with a binding site in the pore domain. The RTG interacts with the pore domain in activated and resting conformations.