Dynamic subunit stoichiometry confers a progressive continuum of pharmacological sensitivity by KCNQ potassium channels

Abstract

Voltage-gated KCNQ1 (Kv7.1) potassium channels are expressed abundantly in heart but they are also found in multiple other tissues. Differential coassembly with single transmembrane KCNE beta subunits in different cell types gives rise to a variety of biophysical properties, hence endowing distinct physiological roles for KCNQ1-KCNEx complexes. Mutations in either KCNQ1 or KCNE1 genes result in diseases in brain, heart, and the respiratory system. In addition to complexities arising from existence of five KCNE subunits, KCNE1 to KCNE5, recent studies in heterologous systems suggest unorthodox stoichiometric dynamics in subunit assembly is dependent on KCNE expression levels. The resultant KCNQ1-KCNE channel complexes may have a range of zero to two or even up to four KCNE subunits coassembling per KCNQ1 tetramer. These findings underscore the need to assess the selectivity of small-molecule KCNQ1 modulators on these different assemblies. Here we report a unique small-molecule gating modulator, ML277, that potentiates both homomultimeric KCNQ1 channels and unsaturated heteromultimeric (KCNQ1)4(KCNE1)n (n < 4) channels. Progressive increase of KCNE1 or KCNE3 expression reduces efficacy of ML277 and eventually abolishes ML277-mediated augmentation. In cardiomyocytes, the slowly activating delayed rectifier potassium current, or IKs, is believed to be a heteromultimeric combination of KCNQ1 and KCNE1, but it is not entirely clear whether IKs is mediated by KCNE-saturated KCNQ1 channels or by channels with intermediate stoichiometries. We found ML277 effectively augments IKs current of cultured human cardiomyocytes and shortens action potential duration. These data indicate that unsaturated heteromultimeric (KCNQ1)4(KCNE1)n channels are present as components of IKs and are pharmacologically distinct from KCNE-saturated KCNQ1-KCNE1 channels.

Keywords: cardiac physiology; drug; long QT; pharmacology.

Fig. 1.

Characterization of ML277 activation of KCNQ1 currents in KCNQ1–CHO cells using manual patch clamp. (A) Representative KCNQ1 current traces in the absence and presence of ML277. (Inset) The recording protocol. The cell was held at −80 mV and depolarizing voltage steps from −70 mV to +50 mV were applied for 2 s in 10-mV increments, followed by a hyperpolarizing step to −120 mV. (B) Time course of the activation effect of ML277 when steady-state currents were measured. (C) Voltage-conductance curves in the absence and presence of ML277, in which the conductance was normalized to the maximal conductance of each treatment (G/G_max), presented as mean ± SEM, n = 4–7. (D) In the presence of ML277, G_max was normalized to the maximal conductance (G_max) in the absence of ML277. The dashed line represents the effect of ML277 after rescaling G_max to 1 as shown in C.

Fig. 2.

Effects of ML277 on inactivation kinetics of KCNQ1. (A) Representative tail current traces to analyze KCNQ1 inactivation kinetics in the absence and presence of ML277 using the protocol shown (Inset). Tail currents were recorded by repolarizing the membrane potential to −120 mV after a series of 2-s depolarizing pulses from −70 mV to +50 mV in 10-mV increments. (B) Tail currents measured at −120 mV after a prepulse to −10, +20 and +50 mV. Tail currents were fitted with a single-exponential function (dashed line) to the later phase of the tail current relaxation. Initial current (x) and extrapolated (y) current are indicated for the currents measured after the prepulse to +50 mV. Horizontal dash line indicates zero current level. (C) Voltage-dependent inactivation curves are shown in the absence and presence of ML277, presented as mean ± SEM (n = 4–7).

Fig. 3.

KCNQ1/KCNE1 complexes lose sensitivity to ML277 when saturated KCNE1 is present. All experiments were completed with transiently transfected cells including WT KCNQ1 alone. (A) Representative KCNQ1/KCNE1 traces for KCNQ1/KCNE1 plasmid ratio 1:0, 1:0.1, and 1:10. (B) Voltage-conductance curves for different ratio of KCNQ1/KCNE1. (C) Bar graph of half-activation voltages for different ratios of KCNQ1/KCNE1. (D) Effect of 0.3 μM ML277 on the different ratios of KCNQ1:KCNE1 (n = 3–7). The KCNQ1 currents were measured at the test voltage +50mV. Data are presented as mean ± SEM.

Fig. 4.

KCNEx subunits reduce the sensitivity of KCNQ1 to activators. (A) A bar graph shows effects of 0.3 μM ML277 on different ratios of KCNQ1 and KCNE subunits (KCNE1 or KCNE3). The ratios of KCNQ1:KCNE1 are 1:0, 1:0.05, 1:0.1, 1:1, and 1:10 from left to right. The ratios of KCNQ1:KCNE3 are 1:0, 1:1, 1:10, 1:50, and 1:100 from left to right. (n = 3–9). (B) KCNQ1/KCNE1 loses sensitivity to R-L3 when saturated KCNE1 dose was present. All experiments were performed with transiently transfected cells including wt KCNQ1 alone. The bar graph shows the effects of R-L3 on the different combinations of KCNQ1:KCNE1 ratios. (n = 3–7). The KCNQ1 currents were measured at the test voltage +50mV. Data are presented as mean ± SEM.

Fig. 5.

ML277 activated unsaturated but not saturated KCNQ1–KCNE1 complexes. All experiments were completed with transiently transfected cells. (A) Representative current traces of EQQ and EQ. The arrows indicate the current level at the depolarization voltage 0 mV (Top and Middle). The superimposed EQ and EQQ current traces at +50 mV to compare the activation time course, in which the current has been normalized to the maximal currents in each cell (Lower). (B) Voltage-conductance curves for different ratios of KCNQ1:KCNE1. EQQ represented KCNQ1:KCNE1 plasmid ratio 1:0.5 and EQ for 1:1 ratio. (C) Bar graph of half-activation voltages for different ratios of KCNQ1:KCNE1. (D) Effect of 0.3 μM ML277 on the different ratios of KCNQ1:KCNE1 (n = 3–7). The KCNQ1 currents were measured at the test voltage +50mV. Data are presented as mean ± SEM. (E) Dose-dependent curves of ML277 on KCNQ1, EQQ, and EQ (n = 4–10).

Fig. 6.

ML277 shortened the action potential duration by increasing potassium currents in iPSC-derived cardiomyocytes. (A) Superimposed representative single action potential traces in the absence and presence of 1 μM ML277. (B) Effect of ML277 on the action potential durations (n = 6). (C) Dose-dependent effect of ML277 on the shortening of action potential duration (APD_90–50) (n = 5–9). (D) Representative current traces for KCNQ1/KCNE1 currents. Current traces (D, a) from transfection molecular ratio 1:0.1 in the CHO-K1 cells and I_Ks currents (D, b) from iPSC-derived cardiomyocytes in the absence and presence of ML277. The recording voltage was same for the two sets of cells. (E) Compound effects on steady-state currents. The holding potential was −60 mV. The steady-state current was examined from −30mV to +50mV in 20-mV increments. Except for I_Ks, other major ionic components including I_Na, I_Kr, I_K1, I_to, and I_CaL were pharmacologically suppressed. ML277-activated currents were chromanol-sensitive. Data are presented as mean ± SEM.