Dual regulation of Kv7.2/7.3 channels by long-chain n-alcohols

Normal alcohols (n-alcohols), also known as primary alcohols, are organic psychoactive substances in which a hydroxyl group (-OH) binds to a primary carbon atom. They consist of short-chain (from one to five carbons, C₁–C₅) and long-chain (C₆–C₂₂) alcohols (Veenstra et al., 2009; Carignan et al., 2013; Ng et al., 2020). Previous studies have revealed that n-alcohols possess anesthetic activity in vivo (Alifimoff et al., 1989; Fang et al., 1997). However, the anesthetic mechanisms of n-alcohols are not fully understood. n-Alcohols have lipophilic properties; thus, they have traditionally been thought to act in a manner that disrupts the cell membrane in the central nervous system (Mullins, 1954; Seeman, 1972; Goldstein, 1984). More recent studies have revealed that alcohols change the ionic movement across the cell membrane by regulating the membrane proteins, including voltage-activated ion channels and ionotropic receptors, in the nervous system (Covarrubias et al., 1995; Dopico and Lovinger, 2009; Garcia et al., 2010; Howard et al., 2014). It has been confirmed that n-alcohols can inhibit sodium channels (Oxford and Swenson, 1979; Klein et al., 2007; Horishita and Harris, 2008), calcium channels (Eckle and Todorovic, 2010; McDavid et al., 2014), Shaw2 Kv channels (Covarrubias et al., 1995; Chu and Treistman, 1997; Shahidullah et al., 2003; Bhattacharji et al., 2010), and Shaker-type Kv channels (Martínez-Morales et al., 2015). In contrast, they can activate the large-conductance, calcium-activated potassium (BK) channels (Chu and Treistman, 1997; Bukiya et al., 2014), and the G protein-coupled inwardly rectifying potassium (GIRK) channels (Bodhinathan and Slesinger, 2013; Bodhinathan and Slesinger, 2014).

The Kv7 channel is a type of voltage-gated potassium channel that is broadly expressed in the central and peripheral nervous systems (Wang et al., 1998; Jentsch, 2000). Of the five members of Kv7 channels, Kv7.2 and Kv7.3 subunits form a stable Kv7.2/7.3 heteromeric channel complex that mediates the muscarine-sensitive M-current (Brown and Adams, 1980). The plasma membrane phosphatidylinositol 4,5-bisphosphate (PtdIns(4,5)P₂) is required for maintaining Kv7.2/7.3 channel activity as a cofactor; hence, the current is suppressed when PtdIns(4,5)P₂ is depleted by G_q-coupled receptor activation (Suh and Hille, 2002; Suh and Hille, 2008; Falkenburger et al., 2010; Telezhkin et al., 2012). Kv7.2/7.3 channels begin to activate from around −40 mV and form non-inactivating currents at −20 mV with slow kinetics for current activation and deactivation in sympathetic ganglion cells (Brown and Adams, 1980; Wang et al., 1998; Brown and Passmore, 2009). Therefore, these channels have been considered to play an important role in determining the membrane potential and excitability of neurons (Greene and Hoshi, 2017).

The anesthetic potency of n-alcohols becomes stronger with an increase in the carbon number in the molecule, and this increase is observed until a cut-off is reached (Fang et al., 1997; Horishita and Harris, 2008; Eckle and Todorovic, 2010). Recently, the effect of n-alcohol was assessed in Kv7.2/7.3 channels using ethanol (EtOH), a short-chain n-alcohol (Kim et al., 2019). The authors reported that EtOH inhibited the M-type Kv7 channels in superior cervical ganglion (SCG) neurons and thus increased neuronal excitability by resetting the resting membrane potential. In the present study, we further assessed the effects of other short-chain and long-chain n-alcohols as well as the EtOH on Kv7.2/7.3 channels. Interestingly, unlike other ion channels showing consistent regulatory tendencies for n-alcohols, we found that Kv7.2/7.3 channels are differentially regulated by n-alcohols depending on their carbon chain length. We further revealed that at least two separate alcohol-binding sites, one for stimulatory effects and the other for inhibitory effects, seem to exist on the Kv7 channel and that the net effect of Kv7 channel stimulation contributes to the anesthetic effects of n-alcohols.

TsA-201 cells were cultured in Dulbecco’s modified Eagle medium (DMEM; HyClone, Thermo Fisher Scientific) supplemented with 10% fetal bovine serum (HyClone, Thermo Fisher Scientific) and 0.2% penicillin/streptomycin (HyClone, Thermo Fisher Scientific) in 100-mm culture dishes at 37°C with 5% CO₂. When the confluency of cells reached about 80%, they were subcultured using Dulbecco’s phosphate-buffered saline (DPBS; Gibco, Thermo Fisher Scientific) to detach the cells. Cells were cultured in 35-mm culture dishes according to the experimental schedule for transfection. For Kv7.2/7.3 channel expression, Kv7.2 (GenBank accession number AF110020) and Kv7.3 (GenBank accession number NM_004519) channels were transiently transfected at a 1:1 M ratio with 200 ng Ds-Red as a transfection marker. Kv7.1 (GenBank accession number NM_000218) which was cloned into the pCMV6-AC-GFP vector was kindly gifted by Dr. Seong Woo Choi (Dongguk University College of Medicine, Seoul, Korea). Lipofectamine 2000 (Invitrogen) was used for transfection when the confluency of cells reached 60–70% in a 35-mm culture dish. The transfected tsA-201 cells were plated onto coverslip chips coated with 0.1 mg/ml poly-L-lysine (Sigma-Aldrich) 36–48 h after transfection. Plated cells were studied in electrophysiological recording 12–24 h after plating.

Sprague-Dawley rats were purchased from SamTako Bio Korea. All animal procedures were approved by the Institutional Animal Care and Use Committee at Daegu Gyeongbuk Institute of Science and Technology (approval number: DGIST-IACUC-20042801-03). Carbon dioxide (CO₂) inhalation was used for the anesthesia of rats. Neonates (5–7 d after birth) were used to prepare SCG.

SCG neurons were cultured using the previously described protocol (Zareen and Greene, 2009). In brief, the isolated ganglia were placed in RPMI 1640 medium (Gibco, Thermo Fisher Scientific), treated with 0.25% trypsin, and incubated for 20 min at 37°C. Following the enzyme treatment, the ganglia were suspended in RPMI 1640 medium containing 10% heat-inactivated horse serum and 5% fetal bovine serum to stop digestion. Then, they were resuspended in RPMI 1640 medium containing 1% heat-inactivated horse serum, 100 ng/ml 2.5 S nerve growth factor, 0.4% penicillin/streptomycin, and 0.4% uridine/5-fluorodeoxyuridine. The cells were plated onto 0.1 mg/l poly-L-lysine with borate buffer chips at 37°C with 5% CO₂. Plated neuronal cells were studied in electrophysiological recording within 2 d after plating.

TsA-201 cells were whole-cell clamped at room temperature (22–25°C) using a HEKA EPC-10 amplifier with Pulse software (HEKA Elektronik). The electrodes were pulled from glass micropipette capillaries using a P-97 micropipette puller (Sutter Instrument) that had a resistance of 2–5 MΩ. The whole-cell access resistance was 2–6 MΩ, and series-resistance errors were compensated by 60%. For recording Kv7 current, cells were held at −20 mV and applied a 500 ms hyperpolarizing step to −60 mV every 4 s. Kv7 currents of SCG neurons were measured in a whole-cell configuration. SCG neuronal cells were held at −20 mV, and a 500 ms hyperpolarizing step was applied to −55 mV every 4 s. Since various ion channels are expressed in SCG neurons, the change in current amplitude at −20 mV can reflect changes in other K⁺ currents as well as Kv7 currents. The Kv7 currents were indirectly obtained from deactivation current recorded at −55 mV as the difference between the average of a 10 ms segment (denoted as “Y0”), taken 10–20 ms into the hyperpolarizing step, and the average during the last 10 ms of that step (indicated as “Y1”). In all voltage protocols, the holding potential was −80 mV.

The extracellular solution used for Kv7 current recording contained 160 mM NaCl, 2.5 mM KCl, 2 mM CaCl₂, 1 mM MgCl₂, 10 mM HEPES, and 8 mM glucose, adjusted to pH 7.4 with NaOH. The intracellular solution of pipette consisted of 175 mM KCl, 5 mM MgCl₂, 5 mM HEPES, 0.1 mM BAPTA, 3 mM Na₂ATP, and 0.1 mM Na₃GTP, adjusted to pH 7.2 with KOH. Retigabine (RTG) was purchased from Sigma-Aldrich. They were stored as 10 mM stocks in DMSO and diluted to working concentrations in extracellular solution on each experimental day. EtOH (purity [GC]: 99.9%) and BuOH (purity [GC]: 99.9%) were purchased from Merck, PrOH (purity [GC]: 99.99%), HeOH (purity [GC]: 99.6%), and OcOH (purity [GC]: 99.7%) used in this study were purchased from Sigma-Aldrich. n-Hexane solution (purity [GC]: 99.4%) and n-octane solution (purity [GC]: 99.4%) were purchased from Alfa Aesar and Merck, respectively. n-Hexane (about 7.65 M) and n-octane (about 6.15 M) were diluted to a working concentration of 10 and 1 mM using K⁺ Ringer’s solution, respectively. We performed a miscibility test and confirmed that the mixture was stable and homogenous without forming a meniscus in the diluted 10 mM n-hexane or 1 mM n-octane over a 24-h.

Single point mutation was introduced into Kv7.2 using the Quikchange II XL site-directed mutagenesis kit (Agilent Technologies). The mutation was verified by DNA sequencing (Macrogen). Primers for cloning were as follows, forward primer (5′→3′): GCT​GGT​CAC​TGC​CTT​GTA​CAT​CGG​CTT​CC and reverse primer (3′→5′): GGA​AGC​CGA​TGT​ACA​AGG​CAG​TGA​CCA​GC.

All data were analyzed using Excel 2016 (Microsoft), IGOR Pro-6.34, and GraphPad Prism 7 (GraphPad Software). Statistics in text or figure were expressed as mean ± SEM. Statistical analysis for differences between the two groups was made by paired or unpaired Student’s t tests. One-way ANOVA was carried out to compare multiple groups. Depending on the variance and sample number, ANOVA was followed by Sidak’s, Games-Howell’s, Tukey’s, or Bonferroni post-hoc test. The time constants were measured by exponential fit. Data for normalized current-voltage (I-V) curves were fitted by the Boltzmann equation. For all analyses, P value < 0.05 was considered significant (*, P < 0.05; **, P < 0.01; and ***, P < 0.001).

Fig. S1 shows the consistent regulation of Kv7.2/7.3 channels by HeOH with different suppliers, grades, and purity. Fig. S2 shows the different modulation of Kv7.2 and Kv7.3 subtypes by OcOH. Fig. S3 shows the effect of HeOH on the activation and deactivation kinetics of Kv7 currents. Fig. S4 shows the insensitivity of mutant Kv7.2(W236L) channels to RTG. Figs. S5 and S6 show supporting data for Fig. 5. Fig. S7 shows the I-V relationship of mutant Kv7.2(W236L) channels insensitive to HeOH and OcOH. Fig. S8 shows the inhibition effect of EtOH on WT and mutant Kv7.2 channels independent of RTG. Fig. S9 shows the dual regulation of long-chain n-alcohols through two independent action sites without allosteric modulation. Fig. S10 shows the electrostatic potential (ESP) surface map of RTG and n-alcohols used in the experiment. Fig. S11 shows the effect of hexane and octane without negative ESP on WT Kv7.2, mutant Kv7.2(W236L) and Kv7.1 channels.

To understand the anesthetic action of n-alcohols, we examined the regulatory effects of various types of n-alcohols on M-type Kv7 currents (I_Kv7) in primary cultured SCG neurons. Short-chain alcohols, such as EtOH, 1-propanol (PrOH), and 1-butanol (BuOH), and long-chain alcohols, such as 1-hexanol (HeOH) and 1-octanol (OcOH), were selected for the analysis (Fig. 1 A, top). Whole-cell patch recording was performed to measure the outward K⁺ currents while applying a depolarization voltage of −20 mV in SCG neurons. The I_Kv7 was calculated by subtracting the average current amplitude at 10 ms before the end of the tail current (Y1) from that at 5–10 ms after the peak of the current (Y0) and is denoted as I_Kv7(Y0-Y1) (Fig. 1 A). Time-course analysis revealed that I_Kv7(Y0-Y1) was inhibited by 31 ± 1% with 200 mM EtOH and by 27 ± 2% with 200 mM PrOH. On the other hand, treatment with 100 mM BuOH, 10 mM HeOH, and 1 mM OcOH resulted in the activation of I_kv7(Y0-Y1) by 2 ± 2, 48 ± 7, and 85 ± 6%, respectively (Fig. 1 C). RTG, the activator of Kv7, except for Kv7.1 (Schrøder et al., 2001; Tatulian et al., 2001; Lange et al., 2009), also enhanced I_kv7(Y0-Y1) in SCG neurons (Fig. 1 B). On using 1 μM of RTG, the concentration within the clinical concentration range (Stas et al., 2016), the Kv7 currents increased by 32 ± 6% (Fig. 1 C).

To further examine how n-alcohols regulate Kv7 channels, the effects of n-alcohols were assessed in human embryonic kidney-derived tsA-201 cells transiently transfected with cDNAs encoding Kv7.2 and Kv7.3. Kv7.2/7.3 currents (I_Kv7.2/7.3) were obtained by applying a voltage of −20 mV. Four different concentrations of n-alcohols were sequentially treated for 40 s each (Fig. 2 A). The representative current traces before and after applying n-alcohols are shown in Fig. 2 B. The results revealed that treatment with 200 mM EtOH and 200 mM PrOH reduced I_Kv7.2/7.3 by 20 ± 2 and 1 ± 2%, respectively. On the other hand, treatment with 100 mM BuOH, 10 mM HeOH, and 1 mM OcOH enhanced the currents by 38 ± 5, 105 ± 7, and 77 ± 6%, respectively. Thus, I_Kv7.2/7.3 was diminished by EtOH but increased by BuOH, HeOH, and OcOH in a concentration-dependent manner (Fig. 2 C). These results were somewhat different from those obtained in SCG neurons. In the neuron, I_kv7(Y0-Y1) was strongly inhibited by PrOH; however, I_Kv7.2/7.3 in tsA-201 cells was not decreased significantly by PrOH. In contrast, BuOH had little effect on I_kv7(Y0-Y1) in SCG neurons, but it significantly increased I_Kv7.2/7.3 in tsA-201 cells.

Many Kv7 channel modulators function in a way that changes the voltage dependence of channel activation (Hou et al., 2019; Zhang et al., 2019; Li et al., 2021b). To further understand the effects of n-alcohols on voltage-dependent activation, we measured the I-V relationship of Kv7.2/7.3 channels before and after treatment with n-alcohols. Fig. 3 A shows the representative current traces elicited in response to test pulses from −70 to +40 mV in 10 mV increments. The overall current amplitudes decreased in the presence of EtOH, did not change in the presence of PrOH, and increased in the presence of BuOH, HeOH, and OcOH. The tail current amplitudes at −80 mV obtained from different voltages were normalized in cells before and after treatment with n-alcohols (Fig. 3 B). We found that 200 mM EtOH did not change the normalized I-V relationship of Kv7.2/7.3 channels (Fig. 3, B and C, top). The application of 200 mM PrOH resulted in a minor change in the I-V relationship. However, the application of 100 mM BuOH, 10 mM HeOH, and 1 mM OcOH shifted the activation curve of Kv7.2/7.3 channels to more negative ranges (Fig. 3, B and C). RTG (10 μM) had little effect on the maximal current amplitude (Li et al., 2021b); however, it induced a strong leftward shift in the I-V relationship (Fig. 3, B and C, bottom). Thus, the results of I_Kv7.2/7.3 elicited in tsA-201 cells indicated that BuOH, HeOH, and OcOH activated the channels by shifting the voltage sensitivity of Kv7.2/7.3 channels to the left, similar to RTG (Li et al., 2021b). On the other hand, EtOH reduced currents without changing the voltage sensitivity of Kv7 channels, while PrOH had no effect on the current amplitude but slightly shifted the voltage-dependent activation.

Among the n-alcohols used in the present study, HeOH activated both I_kv7 in SCG neurons and I_Kv7.2/7.3 in tsA-201 cells. We further investigated how HeOH increases Kv7.2/7.3 channel activity by examining the effects of HeOH on each subtype of Kv7 channels. First, we confirmed that the stimulatory effects of HeOH on current activation and I-V shift of Kv7.2/7.3 were consistent in products from three different commercial sources (Fig. S1), suggesting that the HeOH activity was not due to the chemical impurity or inconsistency in the purification of HeOH during manufacture. On treatment with 10 mM HeOH for 40 s, we observed that the effects differed depending on the Kv7 subtype (Fig. 4). HeOH strongly increased I_Kv7.2 at −20 mV (143 ± 14%); however, I_Kv7.3 was nearly insensitive to HeOH (5 ± 1%). HeOH had intermediate effects on Kv7.2/7.3 heteromeric channels (89 ± 6%; Fig. 4, A and B).

The I-V relationships were also assessed to understand the effect of HeOH on the voltage dependence of channel activation. TsA-201 cells overexpressing Kv7.2, Kv7.3, or both exhibited outward potassium currents in response to test pulses from −70 to +40 mV in 10 mV increments. Representative current traces before and after HeOH treatment are shown in Fig. 4 C. HeOH increased the maximal current amplitudes of Kv7.2 homomeric and Kv7.2/7.3 heteromeric channels; however, the Kv7.3 homomeric channel currents showed no significant change. Unlike the differences in representative current traces among subtypes, the normalized I-V relationship was left-shifted by HeOH in I_Kv7.2, I_Kv7.3, and I_Kv7.2/7.3 (Fig. 4 D). In tsA-201 cells, the half-maximal activation voltage (V_1/2) commonly changed after the application of HeOH in cells expressing Kv7.2, Kv7.2/7.3, and Kv7.3 channels (Fig. 4 E). We also verified that OcOH modulates Kv7 channels with the same tendency as HeOH (Fig. S2).

When HeOH was applied to tsA-201 cells expressing Kv7.3 channels, the current amplitude hardly changed at −20 mV, while the deactivating current at −60 mV dramatically slowed down (Fig. 4 A, right). As RTG has been reported to have a marked effect on I_Kv7 kinetics (Main et al., 2000), we next assessed the effects of HeOH on the kinetics of Kv7 channel activation and deactivation. The representative activation and deactivation traces obtained from tsA-201 cells expressing Kv7.2, heteromeric Kv7.2/7.3, and Kv7.3 channels in the absence or presence of 10 mM HeOH are shown in Fig. S3, A and B. Kv7.2 and Kv7.2/7.3 channels were normally deactivated when applying a voltage step of −60 mV, whereas Kv7.3 channels were not fully deactivated in the presence of HeOH. Thus, a resting potential of −70 mV was used for the activation and deactivation of Kv7.3 channels. In both Kv7.2/7.3 and Kv7.3 channels, treatment with HeOH accelerated the current activation rate and slowed down the deactivation kinetics (Fig. S3 C). In the case of Kv7.2 channels, the activation was mildly accelerated by 10 mM HeOH, while the deactivation was not significantly affected (Fig. S3 C). These results support that HeOH regulates I_Kv7 kinetics, accelerating the rate of channel activation; however, it slows down deactivation kinetics in a subtype-specific manner.

Although not identical, the stimulatory effects of HeOH were reminiscent of RTG, an M-type Kv7 channel activator. Therefore, we used the Kv7.2 homomeric channel to confirm whether n-alcohols can interact with a key residue that is important for the action of RTG. RTG interacts with multiple residues in the Kv7.2 channel (Shi et al., 2020; Li et al., 2021b). Of these, the tryptophan residue W236 is the most essential one. When W236 was substituted with leucine, the Kv7.2(W236L) channels became insensitive to RTG (Schenzer et al., 2005; Wuttke et al., 2005).

All Kv7 channels share a topological design, and four subunits come together to form a single channel, with each subunit consisting of six transmembrane segments (S1–S6; Manville and Abbott, 2018a). Among these, W236 is located in S5 of the Kv7.2 channel. We created a W236L mutant by replacing this residue with leucine (Fig. S4 A). Then, we assessed the current amplitude-dependent activation (I/I_control) at −20 mV to confirm whether the mutant Kv7.2(W236L) channel became insensitive to RTG. The WT Kv7.2 channel was activated by 1, 10, and 30 μM RTG; however, the mutant Kv7.2(W236L) channel was insensitive to all the RTG concentrations (Fig. S4 B).

To examine whether long-chain n-alcohols can activate Kv7.2 channels through W236, the WT Kv7.2 and mutant Kv7.2(W236L) channels were treated with RTG and HeOH. First, the concentration-response curves for RTG and HeOH were determined in WT Kv7.2 channels (Fig. S5). When tsA-201 cells expressing WT Kv7.2 channels were treated with 30 μM RTG, the current amplitude at −20 mV was activated by 86 ± 7%. However, the current amplitude was hardly activated further by 10 mM HeOH in the presence of RTG (Fig. 5, A and B). However, the stimulatory activity of HeOH still appeared in cells pretreated with lower concentrations of RTG (Fig. 5 B). Mutant Kv7.2(W236L) currents were insensitive to 30 μM RTG, while treatment with 10 mM HeOH in the presence of 30 μM RTG induced a significant decrease in the current amplitude (Fig. 5, C and D). As the data could not rule out the possibility that the order of application could affect the results, the compounds were treated in reverse order. Our results showed that the current activation by the second RTG was suppressed by the first HeOH treatment in a dose-dependent manner; HeOH more than 3 mM almost completely inhibited the subsequent current activation by 10 μM RTG (Fig. 5, E and F). Interestingly, when tsA-201 cells expressing the mutant Kv7.2(W236L) channel were treated with HeOH, the current amplitude at −20 mV did not increase; instead, it diminished by 35 ± 6% after treatment of 10 mM HeOH. Treatment with 10 μM RTG in the presence of HeOH did not affect the current amplitude (Fig. 5, G and F). We further assessed the I-V relationship of Kv7.2 channels before and after treatment with 30 μM RTG with or without 10 mM HeOH (Fig. S6 A). Each tail current amplitude at −80 mV obtained from different voltages was normalized (Fig. S6 B). Treatment with 30 μM RTG alone and with 30 μM RTG with 10 mM HeOH shifted the I-V relationship to the left, with no significant difference being noted between the degree of changes in V_1/2 (P = 0.25; Fig. S6 C). Together, these results suggested that the stimulatory effects of HeOH were mediated by left-shifting the I-V curve through interaction with the tryptophan 236 residue in S5, similar to RTG.

We also measured the I-V relationship of W236L channels before and after treatment with HeOH or OcOH to confirm how long-chain n-alcohols modulate V_1/2 on mutant Kv7.2(W236L) channels. The overall current amplitudes decreased in the presence of HeOH or OcOH in W236L channels (Fig. S7 A). The tail current amplitudes at −80 mV obtained from different voltages were normalized in cells before and after treatment with HeOH or OcOH (Fig. S7 B). We found that 10 mM HeOH or 1 mM OcOH did not change the normalized I-V relationship of W236L channels (Fig. S7 C). These results indicate that the voltage-dependency of the current as well as the activation of current amplitude by long-chain n-alcohols were abolished in mutant W236L channels.

Next, we assessed whether the W236 residue is also essential for the inhibitory activity of EtOH. When tsA-201 cells expressing WT Kv7.2 channels were treated with 200 mM EtOH, the amplitude of the currents at −20 mV diminished by 26 ± 2%. WT Kv7.2 currents were activated by 77 ± 13% after treatment with 30 μM RTG, while treatment with 200 mM EtOH in the presence of RTG diminished the currents by 32 ± 4% (Fig. S8, A and C). Mutant Kv7.2(W236L) currents were diminished by 29 ± 1% by 200 mM EtOH; however, they were insensitive to 30 μM RTG (Fig. S8, B and C). The application of 200 mM EtOH in the presence of RTG diminished the currents by 32 ± 3%. No significant differences were noted between the WT Kv7.2 and mutant Kv7.2(W236L) channels in terms of current reduction by EtOH in the absence or presence of RTG (Fig. S8 C). These results indicated that the inhibitory activity of EtOH was independent of W236 in S5.

The functional significant of tryptophan residue in S5 was further examined using Kv7.1 channel in which the S5 tryptophan is innately occupied by a leucine residue (Fig. 6 A). As reported previously (Schenzer et al., 2005), Kv7.1 channel was insensitive to RTG (Fig. 6 B). Similar to Kv7.2/7.3 channels, EtOH inhibited Kv7.1 currents at −20 mV. The long-chain n-alcohols, HeOH and OcOH also reduced the Kv7.1 currents (Fig. 6, B and C). To further identify the effects of n-alcohols on voltage-dependent activation, we measured the I-V relationship of Kv7.1 tail currents at −40 mV before and after treatment with n-alcohols. Fig. 6 D showed the representative current traces elicited in response to test pulses from −80 to +60 mV in 10 mV increments. The overall current amplitudes decreased (Fig. 6 D) and the tail current density at −40 mV was diminished by EtOH, HeOH, and OcOH without shift in the I-V relationships (Fig. 6 E).

To confirm the effect of other n-alcohols, WT Kv7.2 and mutant Kv7.2(W236L) channels were treated with EtOH, PrOH, HeOH, or OcOH. Consistent with the early results, EtOH, PrOH, HeOH, and OcOH differentially regulated the WT Kv7.2 currents (Fig. 7, A and C). In mutant Kv7.2(W236L) channels, the currents were diminished similarly by EtOH, PrOH, HeOH, and OcOH (Fig. 7, B and C). These results suggested the presence of an inhibitory binding site that was independent of the W236 residues and had a similar binding affinity to several types of n-alcohols. In contrast, the conserved tryptophan residue in S5 may mediate the stimulatory effect of long-chain n-alcohols. We thus designed a potential model for studying the effect of n-alcohols on the Kv7 channel (Fig. 7 D).

To further understand the relationship between the stimulatory and inhibitory sites for n-alcohols in Kv7 channels, the effects of EtOH and HeOH were measured in WT Kv7.2 and mutant Kv7.2(W236L) channels in the absence or presence of the activator RTG (Fig. S9). First, the inhibition of Kv7.2 currents (I_Kv7.2) by EtOH was compared in conditions without and with RTG (Fig. S9, A and B). There was no significant difference in percent inhibition of currents by each concentration of EtOH between the two conditions, suggesting that even the stimulatory tryptophan 236 residue was occupied by RTG, the inhibitory effect of EtOH on Kv7.2 channels was not changed. We also verified whether the effects of n-alcohols could be changed by the substitution of tryptophan 236 of the Kv7.2 channel to leucine. As shown in Fig. S9, C and D, since there was no difference in percent inhibition by EtOH between the WT and W236L channels, the inhibitory action by EtOH was not affected by the W236L mutation in Kv7.2 channels. Finally, when the stimulatory and inhibitory effects of HeOH were normalized each other, there was no difference between the two response curves (Fig. S9, E and F). Together, the results suggest that no allosteric modulation exists between the two regulatory sites.

Recent studies have revealed that RTG possesses a negative ESP surface near its carbonyl groups; this ESP is essential for the activation of Kv7.2/7.3 channels (Kim et al., 2015; Shi et al., 2020). We found that the hydroxyl group of n-alcohols has a negative ESP (Fig. S10). To assess whether the hydroxyl group of n-alcohols is also crucial for the modulation of Kv7.2/7.3 channels, we used hexane and octane with the same carbon number as HeOH and OcOH but without any hydroxyl group (Fig. 8, A and G). On treating tsA-201 cells expressing Kv7.2/7.3 channels with 10 mM hexane or 1 mM octane for 40 s, there was no significant change in the current amplitude at −20 mV (Fig. 8, B, C, H, and I). In addition, there were no changes in the I-V relationship and V_1/2 values of Kv7.2/7.3 channels (Fig. 8, D–F and J–L). Similarly, the application of hexane and octane did not affect WT Kv7.2 and mutant Kv7.2(W236L) currents (Fig. S11, A–D). Moreover, the neutral forms did not show the inhibitory effects on Kv7.1 channel (Fig. S11, E and F). This result suggests that the hydroxyl group is required for both stimulatory and inhibitory effects of n-alcohols on Kv7 channels. On the other hand, 3-hexanol with the same number of carbon and hydroxyl group as HeOH but with a different hydroxyl group position, where the negative ESP was located in the middle rather than the end (Fig. 8 M), exhibited only the inhibitory action on Kv7.2/7.3 channels (Fig. 8, N and O), suggesting that position of the ESP surface near the hydroxyl group within the molecule is also important in the regulatory actions of n-alcohols.

In the present study, we found that n-alcohols could differentially regulate I_Kv7.2/7.3 depending on their carbon chain length in SCG neurons. Short-chain n-alcohols, such as EtOH and PrOH, decreased the current amplitude, whereas long-chain n-alcohols, such as HeOH and OcOH, increased the current amplitude. In addition, the longer the carbon chains, the stronger was the potentiation of current amplitudes. We further confirmed that long-chain n-alcohols could shift the voltage dependency of I_Kv7.2/7.3 to more hyperpolarized voltages in tsA-201 cells in a Kv7 subtype-specific manner. Through experiments on mutant Kv7.2(W236L) and Kv7.1 channels with EtOH, HeOH, and RTG, we found that the stimulatory effects of n-alcohols are mediated by a conserved tryptophan residue in S5, while the inhibitory effects of n-alcohols seem to be coupled with the selective EtOH-binding site. Finally, we identified that the essential element of long-chain n-alcohols for Kv7 channel regulation is the hydroxyl head group with a negative ESP. As Kv7.2/7.3 channels play an important role in regulating the resting membrane potential and excitability of SCG neurons (Kim et al., 2019), the present results suggest that long-chain n-alcohols contribute to the anesthetic effects by making it harder to depolarize the membrane and action potentials firing through the net activation of Kv7 channels.

We observed some differences in the regulatory effects of n-alcohols between tsA-201 cells and SCG neurons. This may be due to a couple of cellular properties. The first difference is membrane phospholipid dynamics. Ptdlns(4,5)P₂ is a fundamental signaling cofactor of many ion channels (Hille et al., 2015). It is located in the inner leaflet of plasma membrane and required for many membrane proteins to function as a cofactor. Thus, when the PtdIns(4,5)P₂ is hydrolyzed by phospholipase C (PLC), Kv7 channels are turned off in SCG neurons (Brown and Adams, 1980; Suh and Hille, 2002; Zhang et al., 2003; Winks et al., 2005). According to the previous studies, the resynthesis rate of PtdIns(4,5)P₂ and PtdIns(4)P, a PtdIns(4,5)P₂ precursor, after degradation is much faster in SCG neurons than in tsA-201 cells (Kruse et al., 2016; Kruse and Whitten, 2021). This may cause the difference in the phospholipid composition of the plasma membrane and in the diffusion permeability of n-alcohols through the plasma membrane. The second is ion channel phenotypes expressed in the cells. In our study, tsA-201 cells exhibited outward potassium currents at −20 mV mostly due to overexpression of Kv7.2 and Kv7.3. On the other hand, SCG neurons endogenously express various ion channels, such as TTX-sensitive voltage-gated sodium channels (Rush et al., 2006; Liu et al., 2012), Kv2 channels (Liu and Bean, 2014), and Ca_V2.2 channels (Cheng et al., 2018), as well as Kv7.2 and Kv7.3 (Wang et al., 1998). Although the effects of n-alcohols on those channels are elusive, we speculated that nonspecific regulation of the diverse ion channels by n-alcohols may contribute to the difference in the sensitivity between tsA-201 cells and SCG neurons.

We found that HeOH induces changes in the kinetics of Kv7.2/7.3 channel gating. Although there were differences among subtypes, treatment with HeOH accelerated the current activation rate and slowed down the current deactivation rate. On the other hand, we found that EtOH inhibits I_Kv7.2/7.3 without shifting the channel activation curve. Recent studies by our group have revealed that EtOH suppresses the current by reducing PtdIns(4,5)P₂ sensitivity and the open probability of Kv7.2/7.3 channels instead of controlling a single current or channel number (Kim et al., 2019; Kim and Suh, 2021). Moreover, previous studies have reported that RTG activates Kv7.2/7.3 channels through an increase in open probability (Tatulian and Brown, 2003; Syeda et al., 2016). Therefore, further studies are needed to confirm whether n-alcohols modulate Kv7.2/7.3 channels through the regulation of PtdIns(4,5)P₂ sensitivity and open probability.

In the present study, we found that the chain length of n-alcohols is a crucial factor for Kv7.2/7.3 channel regulation. n-Alcohols consist of a hydrophilic hydroxyl group and hydrophobic carbon chains. As n-alcohols have a single hydroxyl group, their hydrophobicity depends on their chain length (McKarns et al., 1997). EtOH, a short-chain alcohol, has relatively lower hydrophobicity and molar volume (Rottiers et al., 2016); therefore, it can easily pass through the cell membrane and be located on the intracellular side (Patra et al., 2006). On the other hand, long-chain alcohols have high hydrophobicity and are more likely to interact with residues located on the lipid bilayer. Thus, the difference between the polarity and hydrophobicity of n-alcohols leads to different results. Nicotinic acetylcholine receptors have been reported to have at least two separate loci for alcohol actions (Wood et al., 1991). Each site has different chain length requirements—one for activation by short-chain alcohols and the other for inhibition by long-chain alcohols. The present results suggested the possibility that long-chain n-alcohols can bind to two separate working sites on a single Kv7.2/7.3 channel and differently regulate channel gating without allosteric modulation between the sites.

RTG is a neuronal Kv7 channel activator that interacts with several residues in the Kv7 channel. Of these, the conserved tryptophan residue in S5 is the most critical site for activating channel gating (Li et al, 2021b, 2021a). The conserved tryptophan 236 residue in S5 is also a crucial residue for other reported Kv7 activators, such as BMS-204352 (Schrøder et al., 2001; Dupuis et al., 2002; Korsgaard et al., 2005), acrylamide (S)-1 (Bentzen et al., 2006; Jepps et al., 2014), and ML213 (Jepps et al., 2014; Kim et al., 2015). RTG was withdrawn from the market in 2017 because of its low specificity among subtypes, which resulted in several side effects, such as skin discoloration, blurred vision, and urinary retention (French et al., 2011; Brickel et al., 2012), in addition to declined use; however, it is still used for research purposes. In the present study, we found that the mutation of the RTG-binding site in Kv7.2 channels to W236L blocked the activation by HeOH as well as RTG, suggesting that long-chain n-alcohols activate Kv7 channels through the activation mechanism of RTG. This was further confirmed by examining the effects of n-alcohols in Kv7.1 channel which does not have the RTG-binding tryptophan residue in S5. However, it did not change the inhibitory effects of EtOH and HeOH, suggesting that n-alcohols inhibit Kv7 channels through a mechanism independent of the RTG-binding site. Based on research on different potassium channels and EtOH-binding sites (Shahidullah et al., 2003; Bodhinathan and Slesinger, 2014; Bukiya et al., 2014), it may be inferred that the EtOH working site of the Kv7 channel is located in the cytosolic domain. Therefore, the stimulatory and inhibitory sites of n-alcohols are separated in Kv7 channels.

Our results demonstrate that the hydroxyl group (-OH) and its negative ESP is the functional moiety of n-alcohols. The neutral hexane and octane forms showed no activation of Kv7.2/7.3 channels and no inhibition of mutant Kv7.2(W236L) channels, suggesting that the negative ESP is essential for both stimulatory and inhibitory actions of long-chain n-alcohols. Several previous studies also showed that the negative ESP is essential for the regulatory effects of diverse compounds on Kv7 channels. For example, the negative ESP near the carbonyl oxygen moiety was essential for activation of Kv7.2/7.3 channels by a synthetic anticonvulsant, ML-213 (Kim et al., 2015), while it played an important role in the activation of Kv7.2/7.3 channels by glutaconic acid and isovaleric acid (Manville and Abbott, 2018b). Compounds consisting only of hydrocarbons, such as 1-heptene, did not carry negative ESP and did not affect Kv7.2/7.3 currents (Manville and Abbott, 2018b). Since the short-chain n-alcohols without -OH exists as a gas at room temperature, therefore, we could not test the ESP effects of short forms on the currents.

Further research is required to examine whether the changes in I_Kv7.2/7.3 caused by n-alcohols affect action potential generation in neurons and lead to anesthesia in animals. Approximately 10% inhibition of sodium current was found to decrease the propagation of action potential and neural firings (Scholz et al., 1998; Scholz and Vogel, 2000; Horishita and Harris, 2008). On the other hand, ∼25% reduction of M-type Kv7 channel activity by a single mutation led to an abnormal increase in neuronal excitability (Schroeder et al., 1998). However, how much change in Kv7 currents suppresses neuronal firing and induces anesthesia remains unclear. Nevertheless, I_Kv7 was activated by 48 ± 7 and 85 ± 6% on treatment with 10 mM HeOH and 1 mM OcOH, respectively, in SCG neurons. This finding suggests the possibility that long-chain n-alcohols could lead to anesthesia by damping the resting membrane potential and action potential firing. Based on the present results and previous results on sodium channels (Oxford and Swenson, 1979; Klein et al., 2007; Horishita and Harris, 2008), it can be concluded that EtOH and PrOH majorly decrease the propagation of action potentials through sodium channel inhibition, whereas BuOH and other long-chain alcohols induce anesthetic effects through both sodium channel inhibition and Kv7 channel activation.

In summary, the present results suggest that two separate sites interact with n-alcohols: a stimulatory site in the S5 domain and an inhibitory site probably on the intracellular side of the Kv7.2/7.3 channel. Kv7.2/7.3 channels are differently regulated depending on the chain length of n-alcohols. Long-chain n-alcohols have dual effects on separate sites, with the net effect being channel activation. n-Alcohols have been found to have anesthetic properties even before nitrous oxide and diethyl ether were used (Nuland, 1984). However, the molecular mechanism underlying the anesthetic effects remained unclear. The present study demonstrated that long-chain n-alcohols with the negative ESP at position 1 carbon may induce anesthetic effects by increasing Kv7.2/7.3 channel gating through interactions with the conserved stimulatory site tryptophan residue in the S5 domain.

Jeanne M. Nerbonne served as editor.

We thank many laboratories for providing the plasmids.

This work was supported by the National Research Foundation of Korea (NRF) grants funded by the Korean Government Ministry of Sciences and ICT (No. 2022R1A2C100656011 and No. 2020R1A4A1019436).

The authors declare no competing financial interests.

Author contributions: D.J. Jeong and B.C. Suh designed the research; D.J. Jeong and K.W. Kim performed the biochemical and electrophysiological experiments; D.J. Jeong and B.C. Suh wrote the paper; B.C. Suh supervised the project.