Larsson JE , Karlsson U , Wu X , Liin SI .

	Figure 1. Specific endocannabinoids activate the hKV7.2/3 channel expressed in Xenopus oocytes. (a) Molecular structure of 2-AG, AEA, ARA-S, NADA, and NAGABA. (b and c) Mean shift in V50 (ΔV50; b) and mean increase in GMAX (ΔGMAX; c) induced by 10 or 100 µM of indicated endocannabinoid on hKV7.2/3. Data shown as mean ± SEM; n = 4–8. Statistics denote Student’s t test compared with a hypothetical value of 0 with a Bonferroni corrected significance value of P < 0.005. ** indicates P < 0.001. (d) Representative current traces of hKV7.2/3 before and after application of 10 µM ARA-S. Arrow indicates an activating voltage step to −60 mV. Insert of used voltage clamp protocol. (e) Representative G(V) curve for the effect of 10 µM ARA-S for the cell shown in d. Dashed line shows the curve for 10 µM ARA-S normalized to GMAX of control. (f) Concentration-response relation for ARA-S effect on V50. Data shown as mean ± SEM; n = 4–11. ΔV50:MAX (maximal shift in V50) = −32 mV; EC50 = 7 µM. (g) Concentration-response relation for ARA-S effect on GMAX. Data shown as mean ± SEM; n = 4–10. ΔGMAX:MAX (maximal increase of GMAX) = 66%; EC50 = 3 µM. (h) Mean hKV7.2/3 current fold increase at different voltages induced by 10 µM ARA-S. Data shown as mean ± SEM; n = 8.
	Figure S1. Specific endocannabinoids do not activate the hKV7.2/3 channel. (a–c) Representative current traces and corresponding G(V) curves for hKV7.2/3 before and after application of 100 µM indicated endocannabinoids. Arrows indicate an activating voltage step to −40, −60, and −50 mV, respectively. Insert of used voltage clamp protocol.
	Figure S2. Effect of ARA-S and NAGABA on hKV7.2/3. (a and b) Representative example of the wash-in (a) and washout (b) of 10 µM ARA-S with and without supplement of 100 mg/l BSA on hKV7.2/3. The peak current after a 2-s depolarizing step to −40 mV is plotted. Dashed lines indicate basal current level before application of ARA-S. (c) Representative effect of 10 µM ARA-S on the kinetics of hKV7.2/3. Data acquired in high K+ solution (100 mM K+) by activating the channel to +20 mV followed by deactivating pulses from −40 to −130 mV in 10-mV steps. τ, determined by fitting a single exponential to the current generated by the deactivation pulse, was plotted toward the voltage of the deactivation pulse. (d) Representative current traces and corresponding G(V) curves for hKV7.2/3 before and after application of 10 µM NAGABA. Arrows indicate an activating voltage step to −50 mV. Dashed line shows the curve for 10 µM NAGABA normalized to GMAX of control. Insert of used voltage clamp protocol. (e) Concentration-response relation for NAGABA effect on V50. Data shown as mean ± SEM; n = 4–7. ΔV50:MAX (maximal shift in V50) = −21 mV; EC50 = 24 µM. (f) Concentration-response relation for NAGABA effect on GMAX. Data shown as mean ± SEM; n = 3–7. Concentration-response curve could not be converged. (g) Mean hKV7.2/3 current fold increase at different voltages induced by 10 µM NAGABA. Data shown as mean ± SEM; n = 7.
	Figure 2. Importance of the negative charge of the head group and composition of lipid tail for endocannabinoid effect on hKV7.2/3. (a) Representative current traces of hKV7.2/3 before and after application of 10 µM ARA-Serinol. Arrow indicates an activating voltage step to −50 mV. Insert of used voltage clamp protocol and molecular structure of ARA-Serinol. (b) Representative G(V) curve for the effect of 10 µM ARA-Serinol for the cell shown in a. (c) Mean ΔV50 induced by 10 µM ARA-S or 10 µM ARA-Serinol. Data shown as mean ± SEM; n = 4–8. (d) pH response relation for 10 µM ARA-S effect on V50. Data shown as mean ± SEM; n = 3–8. Apparent pKa = 5.4. Inserts indicate the structure of deprotonated and protonated ARA-S head. (e) pH response relation for 10 µM NAGABA effect on V50. Data shown as mean ± SEM; n = 4–7. Apparent pKa = 7.3. Inserts indicate the structure of deprotonated and protonated NAGABA head. (f) Molecular structure of DOC-S, LIN-S, OLE-S, and arachidoyl-S. (g and h) Mean ΔV50 (g) and ΔGMAX (h) induced by 10 µM indicated serine endocannabinoid analogues. Data shown as mean ± SEM; n = 5–9. Statistics denote one-way ANOVA with Dunnett’s multiple comparison test with ARA-S set as control. * indicates P < 0.01; ** indicates P < 0.001.
	Figure S3. Effect of N-arachidonoyl-D-serine on hKV7.2/3. (a) Representative current traces and corresponding G(V) curve of hKV7.2/3 before and after application of 10 µM N-arachidonoyl-D-serine (ARA-D-S). Arrows in the current families indicate an activating voltage step to −50 mV. Insert of used voltage clamp protocol. Dashed line in the G(V) curve shows the curve for 10 µM ARA-D-S normalized to GMAX of control. (b and c) Mean shift in V50 (b) and increase of GMAX (c) induced by 10 µM ARA-S or 10 µM ARA-D-S on hKV7.2/3. Data shown as mean ± SEM; n = 5–8. Statistics denote Student’s t test.
	Figure 3. ARA-S and retigabine activate hKV7.2/3 through different mechanisms. (a) Representative current traces and corresponding G(V) curve of hKV7.2/3 before and after application of 3 µM retigabine (RTG). Arrowheads in the current families indicate an activating voltage step to −50 mV. Insert of used voltage clamp protocol. Dashed line shows the curve for 3 µM retigabine normalized to GMAX of control. (b) Concentration-response relation for retigabine effect on V50. Data shown as mean ± SEM; n = 4–7. ΔV50:MAX (maximal shift in V50) = −35 mV; EC50 = 2 µM. (c) Same as in a, but for hKV7.2_W236L/hKV7.3_W265L. Arrowheads indicate an activating voltage step to −60 mV. (d) Mean shift in V50 induced by 3 µM retigabine on WT hKV7.2/3 or hKV7.2_W236L/hKV7.3_W265L. Data shown as mean ± SEM; n = 4 or 5. Statistics denote Student’s t test. (e) Same as in c, but for the effect of 10 µM ARA-S on hKV7.2_W236L/hKV7.3_W265L. Arrowheads indicate an activating voltage step to −60 mV. (f) Mean shift in V50 induced by 10 µM ARA-S on WT hKV7.2/3 or hKV7.2_W236L/hKV7.3_W265L. Data shown as mean ± SEM; n = 5–8. Statistics denote Student’s t test. ns, nonsignificant; WL, tryptophan to leucine mutation.
	Figure S4. ARA-S activates homomeric hKV7.2 and hKV7.3 tryptophan mutants. (a) Representative current traces and corresponding G(V) curve of hKV7.2_W236L before and after application of 10 µM ARA-S. Arrowheads in the current families indicate an activating voltage step to −60 mV. Insert of used voltage clamp protocol. Dashed line in the G(V) curve shows the curve for 10 µM ARA-S normalized to GMAX of control. (b) Mean shift in V50 induced by 10 µM ARA-S on WT hKV7.2 and hKV7.2_W236L. Data shown as mean ± SEM; n = 6–10. (c and d) Same as in a and b, but for WT hKV7.3_A315T and hKV7.3_W265L_A315T. n = 4–9. Arrowheads in the current families indicate an activating voltage step to −80 mV. (e and f) Same as in a and b, but for WT hKV7.2 and hKV7.2_F168L. n = 6–10. Arrowheads in the current families indicate an activating voltage step to −50 mV. WL, tryptophan to leucine mutation.
	Figure 4. Positively charged residues in S4 and S6 are important for ARA-S activation of hKV7.2 and hKV7.3. (a) Schematic side view of one hKV7 subunit. Transmembrane segments S1–S4, forming the voltage sensor domain, are in gray, and transmembrane segments S5 and S6, forming the pore domain, are in blue. P denotes the pore helix. Amino acid sequences for hKV7.1, hKV7.2, and hKV7.3 are shown above (S4) and below (S6) the channel. Residues mutated to study the ARA-S mechanism of action are indicated in the schematic channel model and sequences. These residues were selected based on previous work identifying residues important for polyunsaturated fatty acid effects on hKV7.1 (Liin et al., 2018). A similar mechanism of action was hypothesized for ARA-S. Top right: the negative charge of the ARA-S head interacts with the first and/or second top arginines of S4 (often called “R1” and “R2,” indicated in red in the S4 sequence) to facilitate the outward movement of S4. S4 movement triggers channel opening. Lower right: In addition, the negative charge of the ARA-S head interacts with the lysine/arginine in the top of S6 (indicated in red in the S6 sequence) to stabilize the selectivity filter in the pore, which increases the overall K+ conductance. Note that hKV7 channels have a glutamine in the natural spot for the third S4 arginine (“Q3,” indicated in green in the S4 sequence). For certain constructs, an arginine was introduced in the Q3 position to restore voltage sensitivity or shift the voltage sensitivity to more WT-like voltages. (b and c) Mean ΔV50 (b) and ΔGMAX (c) induced by 10 µM ARA-S on hKV7.2 with indicated amino acid substitutions. Data shown as mean ± SEM; n = 6–10. Statistics denote one-way ANOVA with Dunnett’s multiple comparison test with WT set as control. The Q204R mutation was introduced to retain voltage sensitivity (see main text for details). nd denotes not determined, as GMAX could not be reliably determined for the R198Q mutant. (d and e) Same as in b and c, but for hKV7.3_A315T with indicated amino acid substitutions (T denotes the substitution of A315T in the constructs). n = 4–9. Statistics denote one-way ANOVA with Dunnett’s multiple comparison test with WTT set as control. The Q233R mutation was introduced to retain voltage sensitivity (see main text for details). nd denotes not determined, as GMAX could not be reliably determined for the R227Q mutant. * indicates P < 0.05; ** indicates P < 0.001.
	Figure S5. Specific charge-neutralizing mutation in hKV7.2 reduces ARA-S effect. Representative current traces and corresponding G(V) curve for indicated mutants before and after application of 10 µM ARA-S. Insert of used voltage clamp protocol. Note the unstable current amplitude of hKV7.2_R198Q, which prevented GMAX determination.
	Figure S6. Specific charge-neutralizing mutation in hKV7.3 reduces ARA-S effect. Representative current traces and corresponding G(V) curve for indicated mutants before and after application of 10 µM ARA-S. Insert of used voltage clamp protocol. Note the unstable current amplitude of hKV7.3_R227Q_A315T, which prevented GMAX determination.
	Figure 5. ARA-S activates all KV7 family members but KV7.4. (a–e) Representative current traces and corresponding G(V) curves for hKV7.1, hKV7.2, hKV7.3_A315T, hKV7.4, and hKV7.5 before and after application of 10 µM ARA-S. Arrowheads indicate an activating voltage step to −60 mV except for KV7.1 and KV7.4, for which the arrowheads indicate an activating voltage step to −50 mV and −30 mV, respectively. Dashed line in the G(V) curve shows the curve for 10 µM ARA-S normalized to GMAX of control. (f and g) Mean shift in V50 (f) and increase in GMAX (g) induced by 10 µM ARA-S. Data shown as mean ± SEM; n = 9–13. The effect of ARA-S on GMAX of hKV7.1 and hKV7.5 might be underestimated because of a tendency of decreasing tail currents in the presence of ARA-S at the most positive voltages (e.g., Fig. 5e), the reason for which remains unknown. (h and i) Same as in f and g, but for 3 µM retigabine (RTG). n = 4–5. Dashed lines in f–i denote effect of indicated treatment on hKV7.2/3.
	Figure 6. Coapplication of low concentrations of ARA-S and retigabine improves the activating effect on hKV7.2/3. (a) Representative current traces and corresponding G(V) curves for hKV7.2/3 before and after coapplication of 1 µM ARA-S and 1 µM retigabine (RTG). Arrowheads indicate an activating voltage step to −60 mV. Insert of used voltage clamp protocols. Dashed line in the G(V) curve shows the curve for coapplication of 1 µM ARA-S and 1 µM retigabine normalized to GMAX of control. (b and c) Mean shift in V50 (b) and increase in GMAX (c) of hKV7.2/3 induced by indicated treatment. Data shown as mean ± SEM; n = 7–10. (d) Mean hKV7.2/3 current fold increase at different voltages induced by indicated treatment. Data shown as mean ± SEM; n = 7–10.
	Figure 7. Coapplication of low concentrations of ARA-S and retigabine limits the off-target effect on other hKV7 subtypes. (a) Mean shift in V50 induced by 10 µM ARA-S, 3 µM retigabine (RTG), or 1 µM ARA-S + 1 µM retigabine coapplied on indicated channels. Data shown as mean ± SEM; n = 4–13. Dashed line denotes a shift in V50 of −20 mV, which was the approximate effect of each treatment on hKV7.2/3. (b) Mean increase in GMAX induced by 10 µM ARA-S, 3 µM retigabine, or 1 µM ARA-S + 1 µM retigabine coapplied on indicated channels. Data shown as mean ± SEM; n = 4–13. Dashed line denotes an increase in GMAX of 50%, which was the approximate effect of 10 µM ARA-S on hKV7.2/3.
	Figure S7. Coapplication of low concentrations of ARA-S and retigabine limits the off-target effect on other hKV7 subtypes. (a and b) Representative current traces and corresponding G(V) curves for hKV7.4 before and after application of either 3 µM retigabine (RTG; a) or coapplication of 1 µM ARA-S and 1 µM retigabine (b). Arrowheads indicate an activating voltage step to −40 mV. Insert of used voltage clamp protocols. (c) Representative current traces and corresponding G(V) curves for hKV7.1 before and after coapplication of 1 µM ARA-S and 1 µM retigabine. Arrowheads indicate an activating voltage step to −40 mV. Insert of used voltage clamp protocols. Dashed lines in the G(V) curves show the curve for each test compound normalized to GMAX of control.
	Figure S8. Coapplication of low concentrations of ARA-S and ICA73 limits the off-target effect on other hKV7 subtypes. (a) Mean shift in V50 induced by 10 µM ARA-S, 20 µM ICA73 (structure shown), or 2 µM ARA-S + 6 µM ICA73 coapplied on indicated channels. Data shown as mean ± SEM; n = 4 or 5. Dashed line denotes a shift in V50 of −20 mV. (b and c) Mean increase in GMAX of hKV7.1 or hKV7.4 induced by indicated treatment. Data shown as mean ± SEM; n = 4 or 5. Statistics denote Student’s t test.

Abdel-Magid, Allosteric modulators: an emerging concept in drug discovery. 2015, Pubmed

Abdel-Magid, Allosteric modulators: an emerging concept in drug discovery. 2015, Pubmed
Bisogno, N-acyl-dopamines: novel synthetic CB(1) cannabinoid-receptor ligands and inhibitors of anandamide inactivation with cannabimimetic activity in vitro and in vivo. 2000, Pubmed
Bohannon, ω-6 and ω-9 polyunsaturated fatty acids with double bonds near the carboxyl head have the highest affinity and largest effects on the cardiac IKs potassium channel. 2019, Pubmed , Xenbase
Bohannon, Polyunsaturated fatty acids produce a range of activators for heterogeneous IKs channel dysfunction. 2020, Pubmed , Xenbase
Bojesen, Binding of anandamide to bovine serum albumin. 2003, Pubmed
Bondarenko, Direct activation of Ca2+ and voltage-gated potassium channels of large conductance by anandamide in endothelial cells does not support the presence of endothelial atypical cannabinoid receptor. 2017, Pubmed
Brodie, Patterns of treatment response in newly diagnosed epilepsy. 2012, Pubmed
Brodie, Efficacy and safety of adjunctive ezogabine (retigabine) in refractory partial epilepsy. 2010, Pubmed
Brueggemann, Differential protein kinase C-dependent modulation of Kv7.4 and Kv7.5 subunits of vascular Kv7 channels. 2014, Pubmed
Börjesson, Lipoelectric modification of ion channel voltage gating by polyunsaturated fatty acids. 2008, Pubmed , Xenbase
Börjesson, An electrostatic potassium channel opener targeting the final voltage sensor transition. 2011, Pubmed , Xenbase
Chadha, Contribution of kv7.4/kv7.5 heteromers to intrinsic and calcitonin gene-related peptide-induced cerebral reactivity. 2014, Pubmed
Changeux, Allosteric Modulation as a Unifying Mechanism for Receptor Function and Regulation. 2016, Pubmed
Choveau, Pore determinants of KCNQ3 K+ current expression. 2012, Pubmed
Choveau, Pore helix-S6 interactions are critical in governing current amplitudes of KCNQ3 K+ channels. 2012, Pubmed
Devane, Isolation and structure of a brain constituent that binds to the cannabinoid receptor. 1992, Pubmed
Di Marzo, The endocannabinoid system and its therapeutic exploitation. 2004, Pubmed
Elinder, Actions and Mechanisms of Polyunsaturated Fatty Acids on Voltage-Gated Ion Channels. 2017, Pubmed
Foster, Allosteric Modulation of GPCRs: New Insights and Potential Utility for Treatment of Schizophrenia and Other CNS Disorders. 2017, Pubmed
French, Randomized, double-blind, placebo-controlled trial of ezogabine (retigabine) in partial epilepsy. 2011, Pubmed
Gantz, Cell-Autonomous Excitation of Midbrain Dopamine Neurons by Endocannabinoid-Dependent Lipid Signaling. 2017, Pubmed
Garin Shkolnik, Blue-gray mucocutaneous discoloration: a new adverse effect of ezogabine. 2014, Pubmed
Godlewski, The endogenous brain constituent N-arachidonoyl L-serine is an activator of large conductance Ca2+-activated K+ channels. 2009, Pubmed
Groseclose, An Investigation into Retigabine (Ezogabine) Associated Dyspigmentation in Rat Eyes by MALDI Imaging Mass Spectrometry. 2019, Pubmed
Gunthorpe, The mechanism of action of retigabine (ezogabine), a first-in-class K+ channel opener for the treatment of epilepsy. 2012, Pubmed
Guo, N-arachidonoyl L-serine, a putative endocannabinoid, alters the activation of N-type Ca2+ channels in sympathetic neurons. 2008, Pubmed
Gómez-Posada, A pore residue of the KCNQ3 potassium M-channel subunit controls surface expression. 2010, Pubmed , Xenbase
Hempel, Metabolism of retigabine (D-23129), a novel anticonvulsant. 1999, Pubmed
Holt, Dominant-negative inhibition of M-like potassium conductances in hair cells of the mouse inner ear. 2007, Pubmed
Huang, Identification of a new class of molecules, the arachidonyl amino acids, and characterization of one member that inhibits pain. 2001, Pubmed
Kim, Atomic basis for therapeutic activation of neuronal potassium channels. 2015, Pubmed , Xenbase
Kondo, 2-Arachidonoylglycerol, an endogenous cannabinoid receptor agonist: identification as one of the major species of monoacylglycerols in various rat tissues, and evidence for its generation through CA2+-dependent and -independent mechanisms. 1998, Pubmed
Kumar, Synthesis and Evaluation of Potent KCNQ2/3-Specific Channel Activators. 2016, Pubmed
Liin, Polyunsaturated fatty acid analogs act antiarrhythmically on the cardiac IKs channel. 2015, Pubmed , Xenbase
Liin, Mechanisms Underlying the Dual Effect of Polyunsaturated Fatty Acid Analogs on Kv7.1. 2018, Pubmed , Xenbase
Liin, Polyunsaturated fatty acids are potent openers of human M-channels expressed in Xenopus laevis oocytes. 2016, Pubmed , Xenbase
Luszczki, Third-generation antiepileptic drugs: mechanisms of action, pharmacokinetics and interactions. 2009, Pubmed
Main, Modulation of KCNQ2/3 potassium channels by the novel anticonvulsant retigabine. 2000, Pubmed , Xenbase
Manville, Ancient and modern anticonvulsants act synergistically in a KCNQ potassium channel binding pocket. 2018, Pubmed , Xenbase
Mechoulam, Identification of an endogenous 2-monoglyceride, present in canine gut, that binds to cannabinoid receptors. 1995, Pubmed
Miceli, Gating consequences of charge neutralization of arginine residues in the S4 segment of K(v)7.2, an epilepsy-linked K+ channel subunit. 2008, Pubmed
Milman, N-arachidonoyl L-serine, an endocannabinoid-like brain constituent with vasodilatory properties. 2006, Pubmed
Panaghie, The role of S4 charges in voltage-dependent and voltage-independent KCNQ1 potassium channel complexes. 2007, Pubmed , Xenbase
Provence, KV7 Channel Pharmacological Activation by the Novel Activator ML213: Role for Heteromeric KV7.4/KV7.5 Channels in Guinea Pig Detrusor Smooth Muscle Function. 2018, Pubmed
Robbins, KCNQ potassium channels: physiology, pathophysiology, and pharmacology. 2001, Pubmed
Rode, Functional effects of the KCNQ modulators retigabine and XE991 in the rat urinary bladder. 2010, Pubmed
Rostock, D-23129: a new anticonvulsant with a broad spectrum activity in animal models of epileptic seizures. 1996, Pubmed
Rundfeldt, The new anticonvulsant retigabine (D-23129) acts as an opener of K+ channels in neuronal cells. 1997, Pubmed
Schenzer, Molecular determinants of KCNQ (Kv7) K+ channel sensitivity to the anticonvulsant retigabine. 2005, Pubmed , Xenbase
Shah, Molecular correlates of the M-current in cultured rat hippocampal neurons. 2002, Pubmed
Silverå Ejneby, Atom-by-atom tuning of the electrostatic potassium-channel modulator dehydroabietic acid. 2018, Pubmed , Xenbase
Splinter, Ezogabine (retigabine) and its role in the treatment of partial-onset seizures: a review. 2012, Pubmed
Tatulian, Activation of expressed KCNQ potassium currents and native neuronal M-type potassium currents by the anti-convulsant drug retigabine. 2001, Pubmed
Taylor, Regulation of KCNQ/Kv7 family voltage-gated K+ channels by lipids. 2017, Pubmed
Tian, Atomic determinants of BK channel activation by polyunsaturated fatty acids. 2016, Pubmed
Tykocki, The KV 7 channel activator retigabine suppresses mouse urinary bladder afferent nerve activity without affecting detrusor smooth muscle K+ channel currents. 2019, Pubmed
Wang, Pore- and voltage sensor-targeted KCNQ openers have distinct state-dependent actions. 2018, Pubmed
Wang, Sequence determinants of subtype-specific actions of KCNQ channel openers. 2017, Pubmed
Wang, KCNQ2 and KCNQ3 potassium channel subunits: molecular correlates of the M-channel. 1998, Pubmed , Xenbase
Wu, KCNE1 remodels the voltage sensor of Kv7.1 to modulate channel function. 2010, Pubmed , Xenbase
Wuttke, The new anticonvulsant retigabine favors voltage-dependent opening of the Kv7.2 (KCNQ2) channel by binding to its activation gate. 2005, Pubmed , Xenbase
Xu, Roles of alternative splicing in the functional properties of inner ear-specific KCNQ4 channels. 2007, Pubmed
Yazdi, The Molecular Basis of Polyunsaturated Fatty Acid Interactions with the Shaker Voltage-Gated Potassium Channel. 2016, Pubmed
Zaydman, PIP2 regulation of KCNQ channels: biophysical and molecular mechanisms for lipid modulation of voltage-dependent gating. 2014, Pubmed
Zhang, KCNQ5/K(v)7.5 potassium channel expression and subcellular localization in primate retinal pigment epithelium and neural retina. 2011, Pubmed