Reversal of Ethanol-induced Intoxication by a Novel Modulator of Gβγ Protein Potentiation of the Glycine Receptor*

The acute intoxicating effects of ethanol in the central nervous system result from the modulation of several molecular targets. It is widely accepted that ethanol enhances the activity of the glycine receptor (GlyR), thus enhancing inhibitory neurotransmission, leading to motor effects, sedation, and respiratory depression. We previously reported that small peptides interfered with the binding of Gβγ to the GlyR and consequently inhibited the ethanol-induced potentiation of the receptor. Now, using virtual screening, we identified a subset of small molecules capable of interacting with the binding site of Gβγ. One of these compounds, M554, inhibited the ethanol potentiation of the GlyR in both evoked currents and synaptic transmission in vitro. When this compound was tested in vivo in mice treated with ethanol (1–3.5 g/kg), it was found to induce a faster recovery of motor incoordination in rotarod experiments and a shorter sedative effect in loss of righting reflex assays. This study describes a novel molecule that might be relevant for the design of useful therapeutic compounds in the treatment of acute alcohol intoxication.

Ethanol is a brain-depressing drug possessing well recognized acute physiological effects. Within the different molecular targets for ethanol, those that best explain the acute effects of ethanol are those affecting fast neurotransmission. It is believed that ethanol effects on sensorial transmission, motor control, respiratory rhythms, and cognitive processing are caused by changes in the activity of several ligand-gated ion channels (1) and particularly through the potentiation of the glycine receptor (GlyR) 2 (2). The GlyR has been extensively studied as a molecular target for ethanol, and electrophysiological experiments have demonstrated that ethanol potentiates GlyR activity (3), as determined by increased glycine-evoked currents (4,5), an increased decay time constant in spontaneous synaptic events (6), increased probability of channel opening in single-channel analysis (7), and increased agonist affinity (8). It was shown previously that G proteins participate in the ethanol effect on the GlyR (4). In addition, more recent studies have determined that the G protein ␤␥ dimer acts as an intermediary for the ethanol action on the GlyR (7,10), where amino acids in the GlyR intracellular domain are essential for the interaction with G␤␥ (11,12). Interestingly, a small peptide (RQHc7) that binds G␤␥ inhibited the potentiation effect of ethanol on both evoked and synaptic currents (6,11). Using in silico analysis, an aspartic pocket (Asp-186, Asp-228, and Asp-246) in G␤ was identified as the binding region for RQHc7. Thus, the inhibition of the interaction between G␤␥ and the GlyR intracellular domain prevented the ethanol effect in native and recombinant systems (6). Therefore, in this study, we aimed to identify small molecules capable of binding G␤␥ and able to inhibit the potentiation of the glycine current induced by ethanol. One of these compounds was assayed in in vivo pharmacology in mice using increasing concentrations of ethanol, demonstrating that it was able to partly reverse some intoxicating effects. Thus, the rationale for designing pharmacological modulators that interfere with the G␤␥-GlyR proteinprotein interaction is amenable for blocking acute ethanol intoxication.

Results
We performed in silico screening aimed at identifying small molecules able to block the effects of ethanol on the GlyR using the region of G␤␥ that was proposed to interact with the GlyR and the RQHc7 peptide (6). The analysis resulted in a group of compounds possessing amide and polar groups that could take part in potential electrostatic and hydrogen bonding interactions with the protein receptor. Several selected compounds contained aromatic and heteroaromatic rings that could be involved in van der Waals andstacking interactions with suitable amino acid side chains. Finally, a subset of 13 compounds was selected to examine their effect on the ethanol * This work was supported by CONICYT FONDECYT Grant 1131004, FONDEF IDeA Grant CA12i10280 (to L. G.), and National Institutes of Health Grant R01-AA15150 (to L. G. A.). The authors declare that they have no conflicts of interest with the contents of this article. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. □ S This article contains supplemental Figure 1 and Table 1 potentiation of the glycine current ( Fig. 1A and supplemental Fig. 1 and Table 1).
The glycine-evoked current at a concentration corresponding to EC 10 (15 M) in HEK293 cells expressing the GlyR␣ 1 subunit is potentiated by 48% Ϯ 3% in the presence of ethanol (100 mM), as shown for the control condition in Fig. 1B (Gly ϩ EtOH) or in the graph of Fig. 1C and as reported previously (7,10,11). The small molecules (200 M) were included in the patch clamp recording pipette to allow them to freely diffuse into the cytoplasm. Recordings of the current with or without ethanol were done immediately after the whole cell configuration was established (minute 1) and after 15 min to allow complete diffusion of the compound. Glycine currents did not have a significant variation between minutes 1 and 15. The ethanol potentiation percentage was obtained from glycinergic currents with and without ethanol at minute 15. None of the molecules included in this study had any effect per se on GlyR activity because glycine-evoked currents (without ethanol) at 1 and 15 min did not present any significant variation. Interestingly, molecules M554, M890, M809a, M194, M363a, M554-1, M554-2, and M890(R) reduced ethanol potentiation, ranging from 19% Ϯ 1% to 33% Ϯ 3% of inhibition (Fig. 1C). In addition, the extracellular application of molecules M554 and M890 inhibited the ethanol potentiation of the glycine current to 29% Ϯ 5% and 33% Ϯ 4%, respectively (Fig. 1D). In this series of experiments, we included M119 because of its well recognized in vivo effect on G␤␥-regulated signaling (13), and as expected, this compound did not induce any effect on the ethanol potentiation of the glycine current (49% Ϯ 3%, Fig. 1C, last column, n Ն 6).
Next, to gain information on the potency of these compounds to inhibit the ethanol potentiation of GlyR activity with M554, M890, M554-1 and M554-2, we performed a detailed pharmacological analysis. For this, increasing concentrations of these compounds were applied, and the effect of ethanol on GlyR-evoked currents was registered ( Fig To obtain data on the predictive intermolecular interaction between G␤␥ and the compounds M554 and M890, studies using molecular docking were done. This analysis provided information on the relative orientation and distances between the chemical groups involved in protein-ligand interactions (Fig. 3A). For example, the indole group of M554 was able to establish an electrostatic ion-dipole interaction with the carboxylate group of Asp-186. Potentially, this indole group could facilitate a van der Waals interaction with Cys-204. On the other hand, the amide group established either a hydrogen bond or an ion-dipole interaction with the carboxylate group of Asp-228 and Asp-246. The urea moiety of M890, similarly, appears to establish a hydrogen bond with the carboxylate group of Asp-228 and also forms an ion-dipole interaction with both Asp-228 and Asp-246. Finally, the hydroxyl group of M890 has the position and distance required to establish a hydrogen bond with Asp-246 (Fig. 3A).
Previous studies have shown that the sites where the diverse effectors in G␤␥ bind are partially overlapped (14). Thus we thought that it was important to determine the existence of functional specificity for our best compounds, i.e. M554 and M890. For this, we first used HL60 cells that express the fMLP receptor coupled to G i to perform a calcium assay linked to the activation of phospholipase C␤ (PLC␤) by G␤␥ (13). These data showed that both M554 and M890 did not have any statistically significant effect on this pathway (Fig. 3B). M119 induce a reduction in intracellular Ca 2ϩ level to 78% of the control.
Together with this, HEK cells expressing the GABA B receptor were used to study the effects of the compounds in the G␤␥ regulation of the G protein-coupled inwardly rectifying potassium channel (GIRK). In this case, M554 and M890 were able to partly inhibit the GIRK activity induced by baclofen by 56% Ϯ 8% and 63% Ϯ 7%, respectively. The compound M119 also had an inhibitory effect in this system of about the same level as M554 (Fig. 3C).
Next we wanted to examine whether these compounds, besides blocking the potentiation of the glycine current by ethanol in GlyR-overexpressing cells, also affected the enhancement of currents caused by activation of the native synaptic GlyR in cultured spinal cord neurons. This was examined by recording miniature inhibitory postsynaptic currents after 15  Interestingly, after application of M554 and M890 (Fig. 4E), ethanol elicited only a very low increment in the decay time constant: 8% Ϯ 5% and 3% Ϯ 7%, respectively. These results show that these molecules also act as inhibitors of the ethanol potentiation in postsynaptical glycinergic events.
Considering the previous results demonstrating the inhibition of ethanol potentiation of glycine currents in recombinant and native GlyRs, we proceeded to examine whether these compounds were also able to attenuate some acute ethanol effects in vivo. Thus, C57BL6/J mice were intraperitoneally administered with ethanol in the absence and presence of the compounds. First, to study locomotor or exploratory activities, we examined the distance traveled in 5 min using an open field assay. After receiving an injection of ethanol (1.0 g/kg), the mice exhibited increased motor activity (p ϭ 0.02541) that was not reversed in the presence of M554 or M890 at a dose of 100 mg/kg (p ϭ 0.34584 and p ϭ 0.81545, respectively; Fig. 5). Mice injected with the compounds alone did not show any change in locomotor activity. To assess the effects of a higher alcohol concentration, motor coordination was examined using an accelerating rotarod assay in the presence and absence of M554 or M890. First, the molecules M554 and M890 injected by themselves did not induce any significant change in motor coordination. The acute administration of ethanol (2.0 g/kg), on the other hand, produced a marked motor incoordination, as described previously (15), with a shorter latency to fall (28% Ϯ 4% of control measured at 15 min post-injection, Fig. 6A). The ethanol-injected mice recovered to 60% Ϯ 6% of pre-injection time, at 105 min. Interestingly, the administration of M554 (100 mg/kg) caused a faster recovery from ethanol-induced motor deficits, reaching 85% Ϯ 8% (compared with 60% Ϯ 6% for the control, p ϭ 0.01329) at minute 75 and 100% Ϯ 8% of recovery at minute 105 after ethanol injection (Fig. 6A). On the other hand, we found no differences in the M890-injected group. M554 did not only antagonize the motor incoordination induced by ethanol, but it also reduced its sedative effects, measured as the time of LORR (Fig. 6B). The data show the effect of ethanol alone and co-administered with different doses of M554 (0, 50, 100, and 200 mg/kg). The injection of 3.5 g/kg ethanol alone produced a sedative effect represented by a LORR of 45 Ϯ 3 min (n ϭ 8). The injection of a 50 mg/kg dose of M554 reduced the LORR to 40 Ϯ 2 min (n ϭ 6, p ϭ 0.2446). Interestingly, increasing the dose to 100 mg/kg reduced LORR to 34 Ϯ 3 min (n ϭ 8, p ϭ 0.04742). No larger effect was found with 200 mg/kg, possibly because of solubility issues. Blood alcohol concentration was determined after 30 min of receiving a dose of 3.5 g/kg of ethanol with or without M554 (100 mg/kg) to discard changes in metabolic clearance that might explain the reduced intoxicating effects of ethanol. The measured blood alcohol concentration values were 3.64 Ϯ 0.29 g/liter for the control (ethanol, n ϭ 3) and 3.84 Ϯ 0.17 g/liter (ethanol plus M554, n ϭ 3), which did not represent any significant difference. Differences in the latency to LORR were not detected in any of the molecules studied (data not shown). Finally, we tested the duration of LORR in KI mice that had a reduced LORR in the presence of ethanol because of point mutations in the intracellular domain of the GlyR␣1 subunit (K385A/K386A GlyR␣ 1 ) important for the intracellular modulation of G␤␥ (16). No differences were found in the duration of LORR with ethanol alone (29 Ϯ 1 min) and ethanol plus M554 (31 Ϯ 1 min, p ϭ 0.27119) in these mice, supporting the notion that the effect of M554 is due to the interference in the intracellular signaling mechanism for the potentiation by ethanol in glycine currents (Fig. 6C).

Discussion
Ethanol is the most commonly abused drug in the world. It is socially accepted at moderate doses because of its anxiolytic and euphoric actions, but it produces serious intoxication at higher doses, leading to loss of motor coordination, sedation, and deep respiratory depression and death. Previous reports have shown the importance of the G␤␥-associated intracellular mechanism for modulation of the GlyR by ethanol and the importance of the intracellular domain in the G␤␥-GlyR interaction (7,12,16). With the aim of developing a pharmacologically active compound, we performed in silico studies followed by neurophysiological experiments and behavioral studies to identify small molecules capable of binding to G␤␥ in the modulatory site for this channel and antagonizing the behavioral actions of ethanol. Supporting this rationale, we were able to identify one compound, named here M554.
Our virtual screening produced a subset of small molecule leads that were examined as inhibitors for the ethanol effects on the GlyR using electrophysiological studies as a first approach to identify active compounds. The strategy was to use the G␤␥ site that binds to the GlyR for the analysis of a small molecule library (6). In this way, the first discrimination parameter was the interaction energy of the molecules to the site in G␤␥. The presence of amine and amide groups in the compounds was expected because of the number of aspartic acid residues in the G␤␥ site used as a receptor in the virtual screening. This type of approach is not unique because other reasonably specific and  potent G␤␥ inhibitors have also been identified following similar approaches (13,17). One of these, M119, was found to bind with good affinity to the G␤ subunit and inhibited PLC␤3 signaling in vitro. Additionally, co-administration of M119 with morphine resulted in a leftward shift in the morphine antinociceptive dose-response curve (13), similar to the shift observed with PLC␤3 knockout mice (18). These and other studies have determined that M119 and its related compound Gallein TM are effective in animal models of inflammation (19), analgesia (20), and heart failure (21). Interestingly, the M119 binding site is similar to the binding site for the SIGK peptide, which is near the binding site of M554. Because M119 was unable to affect the ethanol potentiation of the glycine current, we concluded that the binding site of M119 in G␤␥ does not overlap with the binding site for the GlyR, where M554 and M890 bind. These data are important because the G␤␥ binding sites include those for PLC␤, adenylyl cyclase, and GIRK, among others. M554 and M890 had no effect on PLC␤ activity under the same conditions where M119 was inhibitory for the intracellular Ca 2ϩ increase mediated by this enzyme (13). On the other hand, all assayed compounds, including M119, had inhibitory activity on GIRK channels, possibly because of the critical presence of Asp-228 on the G␤␥ surface (22). The fact that M554 can inhibit a potassium channel is relevant because it might enhance excitability, which is already low in an intoxicated brain. Notably, we did not observe any increase in mouse excitability or behavior, suggesting that the compound is not toxic on its own (Fig. 6A). This is also interesting because future chemical modifications should be expected to improve the selectivity of M554.
The compounds M554 and M890 were also effective at inhibiting the effects of ethanol on synaptic transmission. Glycinergic miniature inhibitory postsynaptic currents were analyzed to evaluate kinetic aspects of the current in the presence and absence of ethanol and to determine their inhibiting effects. In this study, we found that ethanol increased the decay time constant in agreement with previous studies (16,23). Thus, the effect of M554 and M890 of reducing the ethanol enhancement in decay time constant is in agreement with the working model that predicts that the inhibition of the G␤␥-GlyR interaction inhibits the ethanol potentiation of the glycine current.
Behavioral studies have demonstrated that the GlyR is involved in some of the acute effects of ethanol (24,25). Accordingly, intracerebroventricular administration of glycine, or its precursor serine, augmented ethanol-induced LORR in mice (26), and these effects were blocked by strychnine (27). Additionally, KI mice with mutations (Q266I, M287L, or D80A) in the ␣ 1 subunit demonstrated changes in ethanol-induced incoordination and LORR (15,28). Interestingly, it has been shown in our laboratory that a KI mouse with a mutation that reduces ethanol sensitivity of the GlyR␣ 1 subunit reduced ethanol-induced LORR duration (16). Based on this evidence, we assayed different doses of ethanol in animal models, finding that M554 had an inhibitory effect at high ethanol concentrations. M554 did not reverse the enhancing effect of a low dose of ethanol in exploratory activity (Fig. 5), probably because the euphoric effects associated with low concentrations of ethanol are related to modulation of ion channels or proteins that are not regulated by G␤␥ (29). However, when we evaluated motor incoordination using a high dose of ethanol, we found a shorter LORR in the presence of M554, which is in agreement with results indicating the importance of the GlyR in recovery from rotarod ataxia (15). Possible effects of these molecules that could be affecting PLC␤ and GIRK channel activity did not have any influence in behavioral studies because the compounds alone did not show significant changes in both locomotor activity and motor coordination. LORR assays were performed in mice to induce acute intoxication and the impairment of the righting reflex in the presence of M554. The lack of effect of M890 in behavior assays could be determined by an increased renal clearance, liver modification, or for a peripheral distribution, which would be interesting to study in more detail. According to the rationale followed in this work, the ethanol effect induced by the interaction of G␤␥ with the intracellular domain of GlyR is inhibited by the presence of a compound that binds to the same site in G␤␥, acting as a competitive inhibitor of this regulation. Thus, the faster recovery of motor deficits and sedation supports this idea. Indeed, the sedation time induced by ethanol was ϳ25% shorter in the presence of M554. Interestingly, a KI mouse with a mutation in the G␤␥ binding site of the GlyR that reduces ethanol-mediated intracellular modulation of the GlyR␣ 1 subunit (16) showed no difference in duration in ethanol-induced LORR with M554, indicating that the inhibitory effect of M554 is mediated by the GlyR and the signal transduction protein G␤␥.
In summary, a novel, small molecule capable of inhibiting intoxicating ethanol effects in in vitro and in vivo assays by an intracellular mechanism involving G␤␥ as a target was identified following a rational design strategy. To the best of our knowledge, this is the first reported molecule that partially reverses the acute effects of alcohol. Our findings provide information that could be helpful in the design of improved molecules with therapeutic use in the treatment of ethanol-related medical problems.

Experimental Procedures
Virtual Screening and Molecular Docking-A library of molecules was retrieved from the ZINC database (5,548,741 molecules) (9). Property prediction was performed with QikProp (Schrödinger) and filtered with Maestro (Schrödinger), aimed at retaining molecules that satisfy Lipinski and Jorgensen rules, with a Ͼ1LogBB value prediction. As a result of this procedure, 18,691 molecules were selected. After that, high-throughput virtual screening was performed based on Standard Precision docking (GlideSP) and Extra Precision docking (GlideXP) analysis using the three-dimensional structure of G␤␥ taken from the Protein Data Bank (PDB code 1TBG). For the grid receptor generation, the receptor structure of molecular docking was used to generate a 23 ϫ 19 ϫ 23 Å grid containing the Asp-186, Asp-228, and Asp-246 protein residues that have a predicted participation in binding the RQHc7 peptide. A subset of 176 molecules was obtained from this process. Molecular mechanics generalized born solvent accessibility calculation was performed using Prime (Schrödinger) to obtain a preliminary estimation of the binding affinity of the small molecules with the G␤␥ site. Estimated binding free energies were used to select an enriched subset of 13 molecules with highly specific binding and desired properties for further experimental evaluation (supplemental Fig. 1).
Small Molecules-Compounds identified in the virtual screening and used in our experiments were synthetized by the Research Triangle Institute. M119 was obtained from the NCI, National Institutes of Health. The two-dimensional structure of the molecules used in this study was created using Marvin Sketch (ChemAxon, 2013).
Animals-C57BL/6J mice from 9 -14 weeks of age and weighing 20 -30 g were used in behavioral studies. In LORR assays, knockin (KI) mice with a mutation that reduced the intracellular signal modulation of the GlyR␣ 1 subunit were also used (16). All mice were housed 1 week prior to experimentation and had ad libitum access to food and water. Mice were maintained on a 12-h light/dark cycle. Care of the animals and the experimental protocols of this study were approved by the Institutional Animal Use Committee of the University of Concepción and conducted according to the ethical protocols established in the Guide for the Care and Use of Laboratory Animals promulgated by the National Institutes of Health and the National Committee of Science and Technology (CONICYT).
Cell Culture and Transfection-For glycine-evoked current recording, HEK293 cells were transfected with the human GlyR␣ 1 subunit using an Xfect TM transfection reagent kit (Clontech). Expression of GFP was used as a marker of positively transfected cells, and recordings were made after 18 -36 h. For experiments with GIRK channels, HEK293 cells were cultured using standard methodologies and co-transfected with plasmids encoding the GABA B receptor subunits GABA B1 (fused to GFP), GABA B2 , GIRK1, and GIRK4 using an Xfect transfection reagent kit (Clontech).
Electrophysiology-Whole cell recordings were performed using a holding potential of Ϫ60 mV. Patch electrodes were filled with 140 mM KCl, 10 mM 1,2-bis(o-aminophenoxy)ethane-N,N,NЈ,NЈ-tetraacetic acid, 10 mM HEPES (pH 7.4), 4 mM MgCl 2 , 2 mM ATP, and 0.5 mM GTP with or without small molecules at 200 M following the concentration used for the RQHc7 peptide (6). The external solution contained 150 mM NaCl, 5.4 mM KCl, 2.0 mM CaCl 2 , 1.0 mM MgCl 2 , 10 mM HEPES (pH 7.4), and 10 mM glucose. For recording ethanolmediated potentiation of the glycine current, a methodology described previously was used (6). Briefly, glycinergic evoked currents (glycine 15 M) were registered in the presence and absence of ethanol (100 mM) at minutes 1 and 15 after obtaining the whole cell configuration. The maintenance of the glycineevoked current was always checked. Ethanol potentiation was expressed as a percentage between the glycine plus ethanol cur-rent and glycine-only current at minute 15. Isolated glycinergic miniature inhibitory postsynaptic currents from synaptic activity were recorded in the absence and presence of ethanol and molecules (intracellularly applied) after 13-17 days in vitro in spinal neurons. 6-Cyano-7-nitroquinoxaline-2,3-dione (4 M), bicuculline (4 M), and tetrodotoxin (TTX) (100 nM) were added to the external solution to isolate spontaneous glycinergic events. For registering GIRK activity, whole cell recordings were performed in HEK cells expressing the GABA B receptor using a holding potential of Ϫ60 mV. Patch electrodes were filled with 120 mM KCl, 10 mM BAPTA, 10 mM HEPES (pH 7.4), 4 mM MgCl 2 , 2 mM ATP, and 0.5 mM GTP with or without small molecules. The external solution contained 150 mM NaCl, 30 mM KCl, 2.0 mM CaCl 2 , 1.0 mM MgCl 2 , 10 mM HEPES (pH 7.4), and 10 mM glucose. Baclofen was applied in short pulses (4 -5 s) of 10 M every 2 min for 17 min.
Cytosolic Ca 2ϩ Release in Differentiated HL-60 Cells-HL-60 cells expressing the fMLP receptor coupled to G i were seeded in 75-mm 2 flasks at a density of 0.1 ϫ 10 6 cells/ml in 35 ml of RPMI 1640 medium containing 10% FBS, penicillin, and streptomycin. DMSO was added to 1.2%, and cells were allowed to differentiate to neutrophils at 37°C (5% CO 2 ) for 72 h. After differentiation, cells were centrifuged and resuspended in HEPES-buffered saline solution to 250,000 cells/well/100 l. Then cells were loaded with Fluo8 NW for 30 min at 37°C. M554, M890, or M119 was added to the cells for 5 min at room temperature, and then cells were stimulated with formyl-methionyl-leucyl-phenylalanine (fMLP). Intracellular Ca 2ϩ levels were measured using a NOVOstar microplate reader (BMG Labtech).
Open Field Assay-Male and female mice were tested for basal locomotor activity and ethanol-induced locomotor stimulation in a novel open field using a video tracking system (ANY-maze, Stoelting Co.) in C57BL/6J mice. Mice were injected with saline or ethanol (1 g/kg i.p.) in the presence or absence of small molecules 10 min prior to being placed into the 50 ϫ 50 cm test area. Mice were allowed to freely explore the chamber for 5 min, during which time and distance traveled were recorded.
Accelerating Rotarod-Basal motor skill performance was tested in male C57BL/6J mice using an accelerating rotarod assay. Briefly, mice were placed on a non-rotating rod (IITC, Life Science). A timer was started when the rod began accelerating from 4 to 40 rpm in 120 s, and the latency to fall was recorded. Each mouse was tested five times on day 1, day 2, and before injection and four times after injection (every 30 min) on day 3. The latency to fall was normalized for each animal to the time registered immediately before injection. Injections consisted of saline or ethanol (2 g/kg i.p.) in the presence or absence of small molecules.
Loss of Righting Reflex-The sensitivity to ethanol (3.5 g/kg) was determined by using the standard duration of the LORR assay in male C57BL/6J mice (age 9 -13 weeks). Ethanol was diluted in 0.9% saline (20% v/v) and administered i.p. in doses adjusted by injected volumes. Mice were injected with ethanol or ethanol with M554 (100 mg/kg or 200 mg/kg), and when they became ataxic, they were placed in the supine position in V-shaped plastic troughs until they were able to right themselves three times within 30 s. LORR was defined as the time from being placed in the supine position until the mice regained their righting reflex. HP-␤-cyclodextrin was used as a carrier because of the low water solubility of M554. During all LORR assays, normothermia was maintained with the aid of a heat lamp.
Data Analysis-Results are expressed as the mean Ϯ S.E., and statistical analyses were performed using one-way or two-way analysis of variance followed by Bonferroni or Tukey post hoc tests). p Ͻ 0.05 was considered statistically significant. Origin 6.0 (MicroCal) software was used for all statistical analyses.
Author Contributions-L. S. M. conducted most of the electrophysiological experiments, analyzed the results, and contributed to writing the paper. F. C. and V. J. conducted the in silico analysis, analyzed the results, and contributed to writing the paper. C. J. synthetized the small molecules tested and contributed to writing the paper. G. E. Y. contributed to standardizing the behavioral experiments. T. H., D. N., and J. F. conducted the fluorescence measurements. L. G. A. contributed the electrophysiological and behavioral analysis and advice for the project. L. G. conceived the idea for the project, designed the experiments, and wrote most of the paper.