ML252 inhibits Kv7.2 and Kv7.3 homomeric channels with rapid onset. (A to C) Exemplar records from Xenopus oocytes expressing indicated channel types, in (top) 0 µm or (bottom) 10 µm ML252. Oocytes were held at −80 mV and pulsed between −140 and +40 mV in 10 mV steps. (D and E) Oocytes expressing (D) Kv7.2 (n = 7) or (E) Kv7.3* (n = 9) were pulsed from −80 mV to +20 mV for 1 s at 0.1 Hz, and ML252 was applied after the first pulse. (F) Concentration response for ML252 measured at +20 mV from oocytes injected with Kv7.2 (n = 4–13; IC50 = 0.88 μm), Kv7.3 [A315T] (Kv7.3*; n = 8–13; IC50 = 2.71 μm), WT Kv7.3 (Kv7.3; n = 4–8; IC50 = 1.32 μm), Kv7.2/Kv7.3 (n = 3–11; IC50 = 4.05 μm), and Kv7.5 (n = 5–6; IC50 = 6.70 μm). (G) ML252 (black, 10 μm) or XE991 (green, 100 μm) were applied after the first pulse as shown. Oocytes expressing Kv7.2 were pulsed for 1 s at +20 mV, from a holding potential of −80 mV (0.1 Hz). (H) Maximal inhibition of Kv7.2/Kv7.3 heteromers in 100 μm ML252 (black) or XE991 (gray).

Kv7.2 Trp236 strongly influences ML252 inhibition. (A) Docked poses of ML252 (green) in the cryo-EM structure of Kv7.2 (PDB 7VNP), along with the experimentally reported binding mode of ML213 (cyan). The binding pocket is enlarged in the inset, illustrating the orientation of ML252 (left) or ML213 (right) relative to Trp236 (yellow). (B) Exemplar traces from Xenopus oocytes expressing WT Kv7.2 (left) or Kv7.2[W236F] (right), in control or 10 μm ML252, as indicated. Voltage was held at −80 mV and pulsed between −140 and +40 mV (10 mV steps). (C) ML252 concentration response of Kv7.2[W236F] in Xenopus oocytes at +20 mV (n = 5) (WT Kv7.2 data from Figure 1F re-plotted as dashed line). (D and E) Exemplar patch-clamp current recordings of (D) WT Kv7.2 or (E) Kv7.2[W236F] expressed in HEK293 cells. Cells were depolarized and held at +20 mV, followed by drug application or washout where indicated. Segments a and b were used to measure the rate of onset of inhibition, segments c and d were used to measure the rate of recovery. (F) Expanded view of the kinetics of current recovery after ML252 washout in WT Kv7.2 or Kv7.2[W236F] (boxes c and d from panels D and E). (F and G) Traces obtained from ML252 washouts expressing WT Kv7.2 were fit with a Hodgkin–Huxley equation and Kv7.2[W236F] traces were fit with a single exponential function. (G) Rates of ML252 inhibition and recovery in WT Kv7.2 (recovery n = 11, inhibition n = 20) or Kv7.2[W236F] (recovery n = 17, inhibition n = 28) (* indicates P < .001, Student’s t-test). (H) Current magnitudes (normalized to maximal current at +20 mV) in ML252 or after washout, for WT Kv7.2 (n = 12 in 30 μm ML252, n = 8 washout) or Kv7.2[W236F] (n = 25 in 30 μm ML252, n = 15 washout), (* indicates P < .001, Student’s t-test of current inhibition of WT 7.2 vs. 7.2[W236F]).

The pore-targeted activator ML213 weakens ML252 inhibition of Kv7.2. (A) Exemplar records of Kv7.2 expressed in Xenopus oocytes, treated with ML213 and ML252 as indicated. Oocytes were held at −80 mV and pulsed between −140 and +40 mV in 10 mV steps. (B) WT Kv7.2 concentration response to ML252, collected in the presence (open circles, n = 4–6) or absence (dashed line, data from Figure 1F) of ML213. Currents were measured at +20 mV and normalized to current in 0 μm ML252. (C to E) Fast-perfusion patch-clamp recordings using HEK cells showing ML213 competition with ML252. (C) Normalized Kv7.2 current magnitudes from HEK293 cells upon application or washout with indicated combinations of 30 μm ML213 and 30 μm ML252, normalized to currents measured in the absence of drugs and collected as depicted in panels D and E (application, n = 9 for ML213, n = 9 for ML213 + ML252) (washout, n = 4 for ML213, n = 6 for ML213 + ML252). No statistical difference was found for inhibition by ML213 vs. ML213 + ML252 (P = .122, Student’s t-test). (D and E) Exemplar patch-clamp recordings of HEK293 cells expressing WT Kv7.2. Cells were held at +20 mV throughout the recording and combinations of ML252 and ML213 were applied where indicated.

The VSD-targeted activator ICA-069673 does not weaken ML252 inhibition. (A and B) Exemplar records of WT Kv7.2 or Kv7.2[W236F] homomers expressed in Xenopus oocytes, and treated with ICA-069673 (ICA) and ML252 as indicated. Oocytes were held at –80 mV and pulsed between –140 and +40 mV in 10 mV steps. (C) ML252 concentration responses were assessed in WT Kv7.2 and Kv7.2[W236F] channels, in the presence of 30 μm ICA-069673. Currents were recorded at +20 mV and normalized to peak current at +20 mV in ICA-069673 alone (n = 4–7). (D) Normalized WT Kv7.2 and Kv7.2[W236F] current magnitudes in HEK293 cells upon application or washout with indicated combinations of 30 μm ICA-069673 and 30 μm ML252, normalized to currents measured in the absence of drugs and collected as depicted in panels E to H (For WT Kv7.2: ICA-069673 application, n = 9, washout, n = 5; ICA-069673 + ML252 application, n = 12, washout, n = 6) (For Kv7.2[W236F]: ICA-069673 application, n = 12, washout, n = 8; ICA-069673 + ML252 application, n = 12, washout, n = 9 washout of ICA-069673 and ML252). (E to H) Exemplar patch-clamp recordings of HEK293 cells expressing WT Kv7.2 (E and F) or Kv7.2[W236F] (G and H). Cells were held at +20 mV throughout the recording and combinations of ML252 and ICA-069673 were applied where indicated, dashed lines indicate zero current.

Automated patch clamp analysis of ML252 interactions with Kv7 activators in Kv7.2/Kv7.3 heteromeric channels. (A and D) Drug perfusion paradigms to assess functional outcomes of interactions between ML252 and either ML213 or ICA-069673 (ICA-73). In (A), currents were first recorded in control (no drugs added), following a 5-min incubation in ML252 (at ∼IC50 as determined in Figure S4A), and finally after a combination of either activator (multiple concentrations) with ML252. Currents were measured at a pulse to +20 mV where channels are maximally activated, from a holding potential of −80 mV. In (D), currents were first recorded in control (no drugs added), following a 5-min incubation in ICA-73 (30 μm) or ML213 (3 μm), and finally after a combination of either activator with ML252 (multiple concentrations). (B and C) Concentration responses of Kv7.2/Kv7.3 channels to (B) ML213 or (C) ICA-73, administered alone or after preincubation in 1.5 μm ML252. (E and F) Concentration responses of Kv7.2/Kv7.3 to ML252 alone or after preincubation with (E) ML213 or (F) ICA-73. In panels B, C, E, and F, currents (measured at +20 mV) were normalized to response in control conditions in each respective wells containing up to 10 cells (n = 7–8 wells).

ML252 modifies neuronal activity in zebrafish larvae, and is selectively weakened by the pore-targeted activator ML213. (A) Exemplar images of zebrafish larvae (dorsal view) with neuronal-specific expression of CaMPARI. PC (leading to a higher magenta signal) indicates higher neuronal activity, whereas the absence of PC (weaker magenta signal) indicates low neuronal activity. (B) Summary of PC (red/green ratio) observed in zebrafish larvae treated with ML252 alone or in combination with either ML213 or ICA-069673. Intensities of red (R) and green (G) fluorescence emissions are displayed as red to green ratio and normalized to DMSO controls (quantified in the hindbrain area, dashed white line in panel A) (n = 7–9 larvae per condition). * indicates statistical significance vs. ML252 alone (ANOVA followed by Dunn’s post-hoc test, P < .01).

Summary of ML252 inhibition in the context of Kv7 activator binding sites. Mutagenesis and docking suggest that ML252 binds to a similar site as pore-targeted activators including retigabine and ML213 in both Kv7.2 (cyan) and Kv7.3 (red). This underlies less potent ML252-mediated inhibition when applied together with a pore-targeted activator. VSD-targeted activators ztz-240 and ICA-069673 bind to a distinct site (only present in Kv7.2), and therefore do not prevent ML252 inhibition.