Ethanol inhibition of lateral orbitofrontal cortex neuron excitability is mediated via dopamine D1/D5 receptor-induced release of astrocytic glycine

Abstract

Recent findings from this laboratory demonstrate that ethanol reduces the intrinsic excitability of orbitofrontal cortex (OFC) neurons via activation of strychnine-sensitive glycine receptors. Although the mechanism linking ethanol to the release of glycine is currently unknown, astrocytes are a source of neurotransmitters including glycine and activation of dopamine D1-like receptors has been reported to enhance extracellular levels of glycine via a functional reversal of the astrocytic glycine transporter GlyT1. We recently reported that like ethanol, dopamine or a D1/D5 receptor agonist increases a tonic current in lateral OFC (lOFC) neurons. Therefore, in this study, we used whole-cell patch-clamp electrophysiology to examine whether ethanol inhibition of OFC spiking involves the release of glycine from astrocytes and whether this release is dopamine receptor dependent. Ethanol, applied acutely, decreased spiking of lOFC neurons and this effect was blocked by antagonists of GlyT1, the norepinephrine transporter or D1-like but not D2-like receptors. Ethanol enhanced the tonic current of OFC neurons and occluded the effect of dopamine suggesting that ethanol and dopamine may share a common pathway. Altering astrocyte function by suppressing intracellular astrocytic calcium signaling or blocking the astrocyte-specific Kir4.1 potassium channels reduced but did not completely abolish ethanol inhibition of OFC neuron firing. However, when both astrocytic calcium signaling and Kir4.1 channels were inhibited, ethanol had no effect on firing. Ethanol inhibition was also prevented by inhibitors of phospholipase C and conventional isoforms of protein kinase C (cPKC) previously shown to block D1R-induced GlyT1 reversal and PKC inhibition of Kir4.1 channels. Finally, the membrane potential of OFC astrocytes was depolarized by bath application of a Kir4.1 blocker, a D1 agonist or ethanol and ethanol effect was blocked by a D1 antagonist. Together, these findings suggest that acute ethanol inhibits OFC neuron excitability via a D1 receptor-mediated dysregulation of astrocytic glycine transport.

1. Introduction

The orbitofrontal cortex (OFC), a brain region within the prefrontal cortex, has been suggested to play an important role in integrating sensory and reward-related information in support of several aspects of learning such as choice behavior (Roesch and Olson, 2007), reversal learning (Bissonette et al., 2008; McAlonan and Brown, 2003) and the development of anticipation in response to either appetitive or aversive stimuli (Tremblay and Schultz, 1999). Dysfunction of the OFC is associated with numerous neuropsychiatric diseases including alcohol and drug abuse disorders (Fortier et al., 2008; Verdejo-Garcia et al., 2006). We reported previously that ethanol, applied acutely, reduces the intrinsic excitability of OFC neurons in both male and female mice via activation of strychnine-sensitive glycine receptors (Badanich et al., 2013; Nimitvilai et al., 2020). Following chronic intermittent ethanol (CIE) exposure, current-evoked firing of OFC neurons is enhanced for up to two weeks after withdrawal (Nimitvilai et al., 2016), and CIE-treated mice show an impairment in a reversal learning task (Badanich et al., 2011) that in both rodents (Brown et al., 2007) and primates (Jedema et al., 2011) requires the OFC. CIE exposure also diminishes acute ethanol inhibition of spiking by blunting activation of a glycine receptor-dependent tonic current (Nimitvilai et al., 2016). However, the mechanisms that underlie the ethanol-induced release of glycine in the OFC remain unknown.

Recent accumulating evidence has identified astrocytes as critical regulators of neuronal excitability and synaptic transmission. Astrocytes are the most abundant cell type within the central nervous system and perform a variety of functions including providing physical support and nutrients to the neurons, maintaining extracellular ion balance and regulating cerebral blood flow. In addition to these metabolic and structural roles, astrocytes are critical for controlling neuronal excitability and synaptic transmission (Araque et al., 1999). The physical and functional arrangement of the presynaptic terminal, the postsynaptic terminal and the astrocyte known as the tripartite synapse communicate in part via changes in astrocytic Ca²⁺ concentrations. Astrocytic Ca²⁺ signals are primarily mediated through Gq-coupled G-protein receptor-mediated activation of phospholipase C (PLC) and inositol triphosphate (IP₃), that further stimulates Ca²⁺ release from endoplasmic reticulum or Ca²⁺ influx through voltage-gated Ca²⁺ channels (Verkhratsky and Kettenmann, 1996).

Dysfunction of astrocyte activity is implicated in the altered neuronal excitability and neurotransmission observed in alcohol use disorder (AUD) (Butterworth, 1995; de la Monte and Kril, 2014). In addition, blocking glycine receptors interferes with ethanol-induced release of dopamine in the nucleus accumbens (Adermark et al., 2011) suggesting an important role in reward circuits. In a recent study, dopamine was shown to induce the release of glycine from cortical astrocytes via a functional reversal of the glycine transporter 1 (GlyT1) (Shibasaki et al., 2017). In a previous report, we showed that like ethanol, dopamine or a D1/D5 agonist activates a tonic current in lOFC neurons (Nimitvilai et al., 2017), suggesting that ethanol may induce astrocytic glycine release via a dopamine-dependent process. To test this idea, we used multiple approaches to manipulate astrocyte function while recording current-evoked spike firing of OFC neurons or the resting membrane potential of OFC astrocytes.

2. Materials and methods

2.1. Animals

Male C57BL/6J mice were obtained from Jackson Laboratories (Bar Harbor, ME) (https://www.jax.org/strain/00064) at 9 weeks of age. They were group-housed (4/cage) and allowed to acclimatize to the colony room for at least one week in a temperature and humidity controlled AAALAC-approved facility. Animals were maintained on a 12-hour light/dark cycle with lights off at 09:00 am and had ad libitum access to food and water. All animals were treated in strict accordance with the NIH Guide for the Care and Use of Laboratory Animals and all experimental methods were approved by the Medical University of South Carolina’s Institutional Animal Care and Use committee.

2.2. Surgical procedure for in vivo microinjection

To reduce astrocyte Ca²⁺ signaling, an AAV virus encoding the human plasma membrane Ca²⁺ pump 2 (hPMCA2w/b) under control of the astrocyte-specific GfaABC₁D promoter (AAV2/5-GfaABC1D-hPMCA2w/b-mCherry, 350 nl) (Yu et al., 2018) was injected bilaterally into the OFC (AP 2.4, ML+/− 1.35, DV −2.4) of anesthetized mice. For control experiments, an AAV2/5-GfaABC1D-tdTomato virus (350 nl) was injected into the same coordinates of the OFC. Mice were allowed to recover in their home cage for 3–4 weeks before being used in electrophysiology studies.

2.3. Preparation of brain slices

Brain slices containing the lateral orbitofrontal cortex (lOFC) were prepared for whole-cell patch-clamp electrophysiology experiments as previously described (Badanich et al., 2013). Following brief anesthesia with isoflurane and rapid removal of the brain, the tissue was blocked coronally for the frontal cortex; the cerebellum and a portion of the dorsal mesencephalon were removed. The tissue block was mounted in a Leica VT1000S vibratome (Buffalo Grove, IL) containing ice-cold oxygenated (95%O₂, 5%CO₂) sucrose containing buffer and coronal sections (300 μm) were cut. Slices containing the lOFC were held for 30 min in a holding chamber containing 34°C oxygenated artificial cerebral spinal fluid (aCSF) and them maintained at room temperature for at least 30 min before recordings. The composition of the cutting solution used was (in mM): 200 sucrose, 1.9 KCl, 1.2 NaH₂PO₄, 6 MgCl₂, 0.5 CaCl₂, 0.4 ascorbate, 10 glucose, 25 NaHCO₃, adjusted to 305–315 mOsm as needed with sucrose. The composition of the aCSF was (in mM): 125 NaCl, 2.5 KCl, 1.25 NaH₂PO₄, 1.3 MgCl₂, 2.0 CaCl₂, 0.4 ascorbate, 10 glucose, 25 NaHCO₃, adjusted to 290–310 mOsm as needed with sucrose. Both solutions were saturated with 95% O₂/ 5% CO₂ (pH=7.4).

2.4. Whole-cell patch-clamp electrophysiology

An individual slice was placed in the recording chamber and perfused with 34°C aCSF maintained at a flow rate of 2 ml/min. Recordings were localized to deep layers of the lOFC using an Olympus BX51W1 microscope (Center Valley, PA) equipped with infrared Dodt gradient contrast imaging (Luigs and Neumann, Ratingen, Germany) and an infrared camera (Dage-MTI, Michigan City, IN). Thin-wall borosilicate glass electrodes (OD = 1.5 mm, ID = 1.17 mm) were pulled on a Sutter Instrument P97 Micropipette Puller (Novato, CA) and had tip resistances ranging from 2.5–5.2 MΩ. Patch pipettes filled with an internal solution were slowly lowered onto the layer V pyramidal neurons to obtain a high resistance seal (> 1 GOhm) followed by breakthrough to gain whole-cell access. All the whole-cell recordings were carried out using an Axon MultiClamp 700B amplifier (Molecular Devices, Union City, CA) and Instrutech ITC-18 analog-digital converter (HEKA Instruments, Bellmore, NY) controlled by AxographX software (Axograph, Sydney, Australia) running on a Macintosh G4 computer (Apple, Cupertino, CA). Events were filtered at 4kHz and digitized at a sampling rate of 10 kHz.

Current-clamp recordings were performed to determine the effect of ethanol on the intrinsic excitability of OFC neurons in the absence and presence of various modulators. Spike firing was induced by a series of direct current injections (0–220 pA; 750 msec each) through patch pipettes filled with a potassium gluconate internal solution (in mM; 120 K-Gluconate, 10 KCl, 10 HEPES, 2 MgCl₂, 1 EGTA, 2 NaATP, 0.3 NaGTP, adjusted osmolarity to ~294 mOsm with sucrose and adjusted pH to 7.4 with KOH). Changes in tonic current in the absence or presence of acute ethanol and/or dopamine were measured by recording the holding current of a neuron voltage clamped at −70 mV. In these experiments, pipettes were filled with a cesium chloride internal solution (in mM; 120 CsCl, 2 MgCl₂, 1 EGTA, 10 HEPES, and 2NaATP, 0.3 NaGTP, 1 QX-314, adjusted osmolarity to ~294 mOsm with sucrose and adjusted pH to 7.4 with KOH). Drug was locally applied using a gravity-fed perfusion barrel positioned just above the recording area as described previously (Badanich et al., 2013). Specifically, separate glass syringes were used as perfusion reservoirs and a stopcock controlled the flow of each drug solution into a four-way manifold connected to a square quartz perfusion tube (0.6 mm ID; Warner Instruments, Hamden, CT). Before administration, holding current was measured during local perfusion of aCSF to obtain a stable baseline (3 min). Then, the solution was switched to one containing ethanol (66 mM) for 5 min, followed by a combination of dopamine (50 μM) plus ethanol (66 mM) for 5 min. In a separate group of neurons, the change in tonic current produced by local application of dopamine (50 μM) in the absence or presence of the glycine receptor blocker strychnine (1 μM) was measured. Tonic currents were analyzed offline and were measured as changes in current relative to baseline (e.g. I_Ethanol-I_aCSF and I_{Ethanol/Dopamine}-I_Ethanol).

2.5. Sharp electrode recordings

To monitor changes in the membrane potential of astrocytes, OFC slices were treated with the astrocyte-specific fluorescent marker sulforhodamine 101 (0.5 μM SR101, 20 min at 34°C). Following treatment, slices were transferred into regular aCSF and incubated at 34°C for another 10 min to allow washout of excess dye from the extracellular space. Thereafter, slices were kept at room temperature for at least 30 min before recordings. Under current-clamp mode, sharp micropipettes (standard-wall borosilicate glass; OD = 1.5 mm, ID = 0.86 mm) containing 1 M KCl (55–75 MΩ resistance) were lowered onto the SR101-labeled astrocytes until the tip of pipettes contacted the membrane surface. A series of brief voltage pulses (15 V at 10 kHz, 500 μsec duration; Multiclamp Buzz command) were then given to penetrate the astrocyte membrane as indicated by the sudden negative shift in pipette potential. Changes in astrocyte membrane potential were monitored before, during and after bath application of various modulators.

2.6. Drugs

Ethanol was purchased from Pharmco-AAPER (Brookfield, CT). Dopamine hydrochloride, sulpiride, SCH23390, nomifensine, N-[3-([1,1-Biphenyl]-4-yloxy)-3-(4-fluorophenyl)propyl]-N-methylglycine (NFPS), nisoxetine hydrochloride bicuculline, chelerythrine and Gö6976 were purchased from Tocris Biosciences (Minneapolis, MN). The cDNA plasmid pZac2.1-GfaABC1D-hPMCA2w/b-mCherry was purchased from Addgene (plasmid #111568) and used to generate an AAV2/5 virus (Univ. of South Carolina Viral Core, Columbia, SC). The AAV5-GfaABC1D-tdTomato virus was purchased from Addgene. U73122 was purchased from Abcam (Cambridge, MA). Strychnine, VU0134992 hydrochloride, sulforhodamine 101 and all reagents used to prepare aCSF, sucrose-containing and internal pipette solution were purchased from Sigma (St. Louis, MO).

2.7. Statistical analysis

Experimental data are expressed as the mean ± SEM and were analyzed with Prism software (GraphPad Software Inc., San Diego, CA). All current-step spiking data were analyzed using two-way repeated measures ANOVA while total spike number data were analyzed using one-way repeated measures ANOVA. In cases of missing values for repeated measures studies, a mixed model analysis was used and is noted in the text where appropriate. Post hoc tests corrected for multiple comparisons were conducted following all ANOVA analyses as these tests can be performed even if significance in the overall ANOVA is not found (Hsu, 1996; Maxwell et al., 2018). In some experiments, a t-test was used to compare the magnitude of ethanol’s effect under different treatment conditions. Data were considered significantly different when p<0.05.

3. Results

A previous study from this laboratory reported two types of lOFC neurons: large, regular spiking neurons with input resistance <100 MΩ and a small, fast-spiking neurons with higher input resistance and a deep afterhyperpolarization (AHP) amplitude following each action potential (AP) (Badanich et al., 2013). All patch-clamp experiments in the present study were obtained from large, regular-spiking lOFC neurons (N=213). Under control conditions, values recorded for resting membrane potential (−69.78 ± 0.22 mV), input resistance (71.9 ± 1.7 MΩ), AP threshold (−38.31 ± 0.46 mV), AP height (66.21 ± 0.93 mV), rise time (0.35 ± 0.005 ms), half width (1.14 ± 0.02 ms) and AHP amplitude (15.24 ± 0.23 mV) of lOFC neurons were similar to those reported previously by our laboratory (Badanich et al., 2013; Nimitvilai et al., 2016).

3.1. Ethanol inhibition of lateral OFC neuron excitability requires glycine transporter GlyT1.

Similar to that reported previously (Badanich et al., 2013), ethanol (66 mM) significantly reduced the number of current-evoked spikes of lOFC neurons as compared to baseline (Figure 1A; two-way repeated measures ANOVA: main effect of treatment, F_(2,12) = 5.06, p<0.05; Dunnett’s multiple comparison baseline versus ethanol q=2.98, p=0.02, N=7). To provide an overall indication of the magnitude of the ethanol effect, we calculated the total number of spikes evoked by the current step protocol for each condition. As shown in Figure 1B, the number of spikes evoked in the presence of ethanol was significantly less than the baseline (one-way repeated measures ANOVA F_(2,12) = 5.30, p=0.02; Dunnett’s multiple comparison baseline versus ethanol q=3.08, p=0.018). The non-transportable glycine transporter 1 (GlyT1) inhibitor NFPS (0.5 μM) by itself did not alter spiking (Figure 1C). However, in the presence of NFPS, ethanol failed to inhibit firing (two-way repeated measures ANOVA, main effect of treatment F_(2,14) = 0.49, p=0.63; N=8; Figure 1C). Likewise, total spike number before and during NFPS or NFPS+ethanol were not different from one another (one-way repeated measures ANOVA, F_(2,18) = 0.44, p=0.64; Figure 1D). There was a significant difference in the magnitude of inhibition between ethanol alone (47.42 ± 12.93% of the baseline; Fig. 1B) and the ethanol+NFPS group (97.94 ± 9.78% of the baseline, Fig. 1C; one-tailed unpaired t-test, t₍₁₂₎ = 3.12, p=0.0045). Since GlyT1 expression is widely thought to be restricted to glial cells (Adams et al., 1995, but see Cubelos et al., 2005), these results suggest that ethanol inhibition of OFC neuron spiking likely requires functional GlyT1 astrocytic glycine transporters.

Figure 1.

The non-transportable GlyT1 inhibitor NFPS blocks ethanol-induced inhibition of lOFC neuron spiking. Representative traces show decreased action potential (AP) spiking in the presence of 66 mM ethanol as compared with control (scale bars x= 0.1s, y=10 mV). (A and C) Effect of ethanol (66 mM) in the absence or presence of NFPS (0.5 μM) on AP spiking (mean ± SEM) induced by a series of current injections (40–220 pA) in lOFC neurons. (B and D) Total spike number under baseline, NFPS and ethanol plus NFPS conditions. In comparison with control baseline, ethanol produced a significant decrease in basal firing rates of lOFC neurons (A, two-way repeated measures ANOVA: main effect of ethanol, F_(2,12) = 5.06, p<0.05; Dunnett’s multiple comparison baseline versus ethanol q=2.98, p=0.02, N=7, 3 mice). Total spike number during ethanol application was also significantly different from those from baseline (B, one-way repeated measures ANOVA F_(2,12) = 5.30, p=0.02; Dunnett’s multiple comparison baseline versus ethanol q=3.08, p=0.018). NFPS (0.5 μM) alone did not alter spiking firing but blocks ethanol inhibition of OFC firing; in that no difference in spike number was observed (C, two-way repeated measures ANOVA, main effect of treatment F_(2,14) = 0.49, p=0.63; N=8, 3 mice). Total spike numbers before and during NFPS or NFPS+ethanol application were not significantly different (D, one-way repeated measures ANOVA, F_(2,14) = 0.46, p=0.64). Symbols: (*) p<0.05

3.2. Ethanol inhibition of lateral OFC neuron excitability requires norepinephrine transporters.

We reported previously that ethanol reduces the intrinsic excitability of OFC neurons via increased tonic glycine current (Badanich et al., 2013). Like ethanol, dopamine or the D1 agonist SKF81297 also increases tonic current and this increase is blocked by the D1-like receptor antagonist SCH23390 or picrotoxin (Nimitvilai et al., 2017) In this study, we repeated those experiments and measured the holding current before and during local application of dopamine, the glycine receptor inhibitor strychnine or dopamine plus strychnine. Dopamine (50 μM) produced a negative shift in the holding current of lOFC neurons (−33.48 ± 7.29 pA) that was significantly different from the pre-dopamine baseline (two-tailed paired t-test, t₍₁₅₎ = 4.60, p=0.0003; N=16). The mean holding currents before and during dopamine application were − 228.3 ± 36.61 and −261.8 ± 36.92 pA, respectively. Application of strychnine (1 μM) or dopamine+strychnine produced a small positive shift in the holding currents (8.64 ± 11.77 pA for strychnine; 11.04 ± 5.7 pA for dopamine + strychnine) that was not different from control baseline (one-way repeated measures ANOVA, main treatment effect, F_(2,14) = 1.09, p=0.37; N=8). Post-hoc comparisons found no significant difference in currents for strychnine (q=0.64, p>0.05) or dopamine+strychnine conditions (q=2.03, p>0.05). Likewise, there was no difference in the holding current of lOFC neurons in the presence of dopamine+strychnine as compared to the strychnine baseline (q=1.43, p>0.05). The mean holding currents before and during strychnine and strychnine+dopamine were −230.0 ± 26.96, −225.4 ± 25.73 and −214.4 ± 27.27 pA, respectively. These results suggest that dopamine increases the tonic current of lOFC neurons via activation of strychnine-sensitive glycine receptor.

A recent study from this laboratory demonstrated that extracellular levels of dopamine in cortical slices are regulated by monoamine transporters (Nimitvilai et al., 2017). Dopamine has been shown to induce the release of glycine from cortical astrocytes via a functional reversal of the glycine transporter 1 (GlyT1) (Shibasaki et al., 2017). Since both ethanol and dopamine increase glycine-mediated tonic currents of lOFC neurons, we examined whether blocking monoamine transporters would interfere with ethanol inhibition of spiking. In initial experiments, baseline recordings were performed followed by bath application of nomifensine (10 μM) and then nomifensine plus ethanol (66 mM) As shown in Figure 2A and B, there was no effect of nomifensine by itself on spike firing or total spikes but it blocked the inhibition of firing normally observed with ethanol (Figure 2A; two-way repeated measures ANOVA: main effect of treatment F_(2,20) = 0.0015, p=0.99; N=10; Figure 2B; one-way repeated measures mixed model, F_(2,18) = 0.44, p=0.65). In the frontal cortex, expression of dopamine transporters is sparse (Freed et al., 1995; Morón et al., 2002; Sesack et al., 1998) and dopamine has been shown to be taken up by the norepinephrine transporters (NET) (Morón et al., 2002; Mundorf et al., 2001; Yamamoto and Novotney, 1998). Therefore, we tested whether the selective NET blocker nisoxetine would also prevent ethanol inhibition of OFC neuron spiking. Recordings were carried out under baseline conditions and in the presence of nisoxetine (50 nM), nisoxetine plus ethanol (66 mM) followed by a washout. Two-way repeated measures ANOVA revealed a significant main effect of treatment (Figure 2C; F_(3,39)=10.42, p<0.0001) that was driven by a difference in spiking between the baseline and nisoxetine conditions (Sidak’s multiple comparison baseline versus nisoxetine t=3.45, p=0.008), nisoxetine plus ethanol (Sidak’s multiple comparison t=5.53, p<0.0001) and the washout (Sidak’s multiple comparison baseline versus washout t=3.16, p=0.018). However, there was no difference in spiking between the nisoxetine and nisoxetine plus ethanol groups (Sidak’s multiple comparison t=2.08, p=0.239). When total spike numbers were compared, there was again a significant main effect (Figure 2D; one-way repeated measures mixed model, main effect of treatment F_(3,39)=10.42, p<0.0001) that was driven by a difference between the baseline and nisoxetine conditions (Sidak’s multiple comparison t=3.45, p=0.008), nisoxetine plus ethanol (Sidak’s multiple comparison t=5.53, p<0.0001) and the washout (Sidak’s multiple comparison baseline versus washout t=3.16, p=0.02). There was no difference in total spikes between the nisoxetine and nisoxetine plus ethanol groups (Sidak’s multiple comparison t=2.08, p=0.24). These findings suggest that in the OFC, ethanol may cause reversal of a norepinephrine transporter resulting in release of dopamine.

Figure 2.

Ethanol-induced inhibition of OFC spiking is blocked by the norepinephrine transporter inhibitor. (A and C) Effect of ethanol (66 mM) in the absence or presence nomifensine (10 μM) or nisoxetine (10 μM) on AP spiking (mean ± SEM) induced by a series of current injections (40–220 pA) in lOFC neurons. (B and D) Total spike number under baseline, nomifensine or nisoxetine and ethanol plus nomifensine/nisoxetine conditions. (A) In the presence of the nonspecific monoamine transporter inhibitor nomifensine, ethanol did not alter action potential spiking of OFC neurons (two-way repeated measures ANOVA, F_(2,20) = 0.0015, p=0.99; N=10, 5 mice). (B) Total spike numbers at baseline and during nomifensine and ethanol+nomifensine application were also not different (one-way repeated measures mixed model, F_(2,18) = 0.44, p=0.65). (C) Although the NET blocker nisoxetine (50 nM) by itself significantly decreased spike firing of lOFC neurons (two-way repeated measures ANOVA, F_(3,39)=10.42, p<0.0001; Dunnett’s multiple comparisons baseline versus nisoxetine q=3.45, p=0.004), there was no change in spiking by ethanol in the presence of nisoxetine (Dunnett’s multiple comparison q=2.08, p=0.11, N=16, 5 mice). (D) There was a main effect of treatment on total spike numbers (one-way repeated measures mixed model, main effect of treatment F_(3,39)=10.50, p<0.0001; Dunnett’s multiple comparison, baseline versus nisoxetine alone q=3.41, p=0.004) but no difference between the nisoxetine and nisoxetine plus ethanol groups (q-2.15, p=0.10). Symbol: (**) p<0.01.

3.3. Ethanol inhibition of lateral OFC neuron excitability involves dopamine D1/D5 receptors.

Dopamine-induced enhancement of tonic currents is blocked by a D1/D5 antagonist but not a D2 antagonist (Nimitvilai et al., 2017), and both ethanol and dopamine increase glycine-mediated tonic currents of lOFC neurons. Since blocking dopamine uptake by the NET antagonist nisoxetine suppressed ethanol inhibition of OFC spiking, we tested whether ethanol inhibition of intrinsic excitability is dopamine receptor dependent using a within-neuron recording protocol. Current-evoked spiking was first measured under baseline conditions followed by application of ethanol and a subsequent washout until obtaining a second stable baseline. Then, the D2 antagonist sulpiride (10 μM) or the D1/D5 antagonist SCH23390 (10 μM) was applied for 10 min, followed by perfusion with ethanol plus sulpiride or SCH23390 for another 8 min. Under control conditions in the sulpiride experiment, ethanol (66 mM) significantly inhibited spike firing of OFC neurons as compared to the pre-ethanol baseline (Figure 3A; two-way repeated measures ANOVA: main effect of treatment, F_(2,16) = 16.49, p=0.0001; Dunnett’s multiple comparison baseline versus ethanol, q=5.27, p=0.0002). Total spike number was also significantly reduced by ethanol (Figure 3C; two-way repeated measures ANOVA main effect of ethanol F_(1,16)=28.00, p<0.0001; Sidak’s multiple comparison baseline versus ethanol t=4.95, p<0.0003) and was 49.39 ± 10.75% of the control value. In the presence of the D2 antagonist sulpiride (10 μM), that by itself had no effect, ethanol still decreased lOFC neuron firing (Figure 3B; two-way repeated measures ANOVA main effect of treatment, F_(2,16)=6.47, p=0.0087; Dunnett’s multiple comparison sulpiride versus sulpiride plus ethanol q=3.28, p=0.009). To determine whether the magnitude of ethanol inhibition was different in the absence and presence of sulpiride, we first calculated the total number of spikes generated by the current steps under each condition and then used this value to determine ethanol inhibition. As shown in Figure 3C, ethanol inhibited spike firing under control conditions by 50.61 ± 10.75% and by 37.77 ± 13.09% in the presence of sulpiride and these values were not different from one another (one-tailed paired t-test, t₍₈₎=1.34, p=0.11). In the control recordings of the D1/D5 antagonist SCH23390 experiment, ethanol (66 mM) produced a significant decrease in AP firing as compared to the pre-ethanol baseline (Figure 3D; two-way repeated measures ANOVA: main effect of ethanol, F_(2,18) = 12.39, p<0.0004; N=10). There was a main effect of treatment when recordings were carried out in the presence of SCH23390 (Figure 3E; two-way repeated measures ANOVA _F(2,18)=6.01, p=0.01) but this was due to a significant difference between the second baseline and the SCH23390 plus ethanol condition (Dunnett’s multiple comparison wash ethanol versus SCH23990 plus ethanol q=3.46, p=0.005). There was no difference in the current-evoked firing curves between the SCH23390 and SCH23990 plus ethanol groups (Dunnett’s multiple comparison SCH23390 versus SCH23990 plus ethanol q=2.08, p=0.093). Figure 3F compares the magnitude of ethanol inhibition under control and SCH23390 conditions using the total spike number as described above. In the absence of SCH23390, ethanol inhibited spike firing by 39.21 ± 10.29% while inhibition was 15.72 ± 5.48% in the presence of SCH23390 and these values were significantly different from one another (one-tailed paired t-test t₍₈₎=2.24, p=0.028). These data suggest that D1/D5 receptor activity enhances the ethanol inhibition of OFC firing.

Figure 3.

Ethanol-induced inhibition of OFC spiking is blocked by the D1/D5, but not D2 receptor antagonist, and by the norepinephrine transporter inhibitor. Effect of ethanol (66 mM) in the absence (A and D) or presence of sulpiride (B) or SCH23390 (E) on AP spiking (mean ± SEM) induced by a series of current injections (40–220 pA) in lOFC neurons. (A) As compared to baseline, ethanol (66 mM) significantly decreased AP spiking of OFC neurons (two-way repeated measures ANOVA: main effect of ethanol, F_(2,16) = 16.49, p=0.0001; Dunnett’s multiple comparison baseline versus ethanol, q=5.27, p=0.0002; N=9, 5 mice). (B) Ethanol still decreased lOFC neuron firing in the presence of sulpiride (10 μM, two-way repeated measures ANOVA main effect of treatment, F_(2,16)=6.47, p=0.0087; Dunnett’s multiple comparison sulpiride versus sulpiride plus ethanol q=3.28, p=0.009) that by itself had no effect (p>0.05). (C) The magnitude of ethanol inhibition calculated from the total spike number was not different in the absence and presence of sulpiride (one-tailed paired t-test, t=1.42, p=0.099). D) Prior to application of SCH23390, ethanol produced a significant decrease in AP firing (two-way repeated measures ANOVA: main effect of ethanol, F_(2,18) = 12.39, p<0.0004; N=10, 5 mice). (E) There was a main effect when recordings were carried out in the presence of the D1/D5 antagonist SCH23390 (10 μM; two-way repeated measures ANOVA _F(2,18)=6.01, p=0.01). This was due to a significant difference between the second baseline and the SCH23390 plus ethanol condition (Dunnett’s multiple comparison wash ethanol versus SCH23990 plus ethanol q=3.46, p=0.005) as there was no difference in the current-evoked firing curves between the SCH23390 and SCH23990 plus ethanol groups (Dunnett’s multiple comparison SCH23390 versus SCH23990 plus ethanol q=2.08, p=0.093). (F) The magnitude of ethanol inhibition calculated from the total spike number was significantly reduced by SCH23390 (one-tailed paired t-test, t=2.24, p=0.028). Symbols: (*) p<0.05, (**) p<0.01, (***) p<0.001.

3.4. Over-expression of a Ca²⁺ pump in astrocytes together with a Kir4.1 channel inhibitor suppresses ethanol inhibition of lateral OFC neuron excitability

Many studies have identified astrocytes as a source of neuromodulators including glycine (Roux and Supplisson, 2000; Sakata et al., 1997). Under physiological conditions, GlyT1 takes up excess glycine from synaptic sites, accompanied by co-transport of 2Na⁺ and Cl⁻ (Huang et al., 2004). However, the astrocytic GlyT1 can reverse and release glycine when intracellular concentrations of glycine are increased (Huang et al., 2004) or when the astrocyte membrane potential is depolarized (Roux et al., 2000; Sakata et al., 1997). In addition, dopamine has been reported to cause a functional reversal of the astrocytic GlyT1 transporter resulting in elevated extracellular levels of glycine (Shibasaki et al., 2017). Regulation of astrocytic intracellular calcium has also been shown to be important for proper neuronal function (Bazargani and Attwell, 2016) and dysfunction of astrocyte Ca²⁺ signaling results in behavioral deficits and neurological disease (Khakh and Sofroniew, 2015; Yu et al., 2018). We thus explored whether dampening astrocytic intracellular calcium would interfere with the ethanol inhibition of OFC neuron firing. In these experiments, mice were injected with astrocyte-selective AAVs encoding a fluorescent reporter (GfaABC1D-tdTomato) or the human plasma membrane Ca²⁺ pump PMCA (GfaABC1D-mCherry-hPMCA2w/b; Figure 4A) previously shown to blunt astrocyte calcium signaling (Yu et al., 2018). Under baseline conditions, there were no changes in current-evoked firing in OFC neurons from mice expressing hPMCA2w/b or tdTomato in astrocytes as compared to those from naïve mice (one-way repeated measures mixed model F_(2,32)=0.934, p=0.403; Sidak’s multiple comparison naïve mice versus hPMCA2w/b-injected mice t=0.75, p=0.84, naïve mice versus tdTomato mice t=1.27, p=0.51, hPMCA2w/b versus tdTomato mice t=0.368, p=0.977). In mice injected with the AAV-GfaABC1D-tdTomato virus, ethanol produced a significant reduction in AP spiking of OFC neurons (Figure 4B; two-way repeated measures ANOVA: main effect of treatment, F_(2,34) = 30.26, p<0.0001; Dunnett’s multiple comparison baseline versus ethanol q=6.78, p<0.0001 N=18). Likewise, total spike number during ethanol application to tdTomato expressing slices was significantly different from that of baseline (Figure 4C; one-way repeated measures ANOVA, F_(2,34) = 30.26, p<0.0001; Dunnett’s multiple comparison baseline versus ethanol q=6.78, p<0.0001). In neurons from hPMCA2w/b expressing animals, ethanol produced a small but statistically significant decrease in OFC firing as compared to baseline (Figure 4D; two-way repeated measures ANOVA: main effect of treatment, F_(2,30) = 4.07, p<0.03; Dunnett’s multiple comparison baseline versus ethanol q=2.42, p=0.04, N=16). Total spike number during ethanol application from hPMCA2w/b expressing mice also showed a small but significant decrease as compared to baseline (Figure 4E; one-way repeated measures ANOVA, F_(2,27) = 4.07, p=0.029). The ethanol-induced change in lOFC spiking in slices from the hPMCA2w/b group (81.96 ± 10.11% of baseline, Fig. 4E) was significantly less than that of the tdTomato control group (47.39 ± 8.17% of baseline, Fig 4C; one-tailed unpaired t-test t₍₃₂₎ = 2.68, p=0.0057). This result suggests that while actively buffering astrocytic Ca²⁺ signaling significantly blunts ethanol inhibition of spiking, other processes may be involved or a more complete inhibition of astrocyte activity is required to fully suppress ethanol’s effect.

Figure 4.

Over-expression of a Ca²⁺ pump in astrocytes reduces ethanol-induced inhibition of lOFC neuron spiking. (A) Left image shows expression of AAV-GfaABC1D-tdTomato in the lateral OFC (2X magnification). Middle image shows lOFC astrocytes from a mouse injected with AAV-GfaABC1D-mCherry-hPMCA2w/b (40X magnification). Right panel is an image of live lOFC astrocytes (40X magnification) taken during a slice recording from a mouse injected with AAV-GfaABC1D-mCherry-hPMCA2w/b. (B) In slices from AAV-GfaABC1D-tdTomato control animals, ethanol (66 mM) significantly reduced AP spiking of lOFC neurons (two-way repeated measures ANOVA: main effect of treatment, F_(2,34) = 30.26, p<0.0001; Dunnett’s multiple comparison baseline versus ethanol q=6.78, p<0.0001 N=18, 3 mice). (C) Total spike number during ethanol application was also significantly different from those from baseline (one-way repeated measures ANOVA, F_(2,34) = 30.26, p<0.0001; Dunnett’s multiple comparison baseline versus ethanol q=6.78, p<0.0001). (D) In slices from hPMCA2w/b injected mice, ethanol produced a small but significant decrease in basal firing rates of lOFC neurons (two-way repeated measures ANOVA: main effect of treatment, F_(2,30) = 4.07, p<0.03; Dunnett’s multiple comparison baseline versus ethanol q=2.42, p=0.04) N=16, 5 mice). (E) Total spike number during ethanol application from hPMCA2w/b expressing mice also showed a small but significant difference from baseline (one-way repeated measures ANOVA, F_(2,27) = 4.07, p=0.029). Symbols: (*) p<0.05, (****) p<0.0001.

Besides removing neurotransmitters from synapses, astrocytes are critical for regulating extracellular levels of potassium. Inwardly rectifying Kir4.1 channels are the main K⁺ conductance in astrocytes and are not expressed in neurons (Brasko et al., 2017; Butt and Kalsi, 2006; Higashi et al., 2001). Pharmacologically blocking Kir4.1 channels or genetically deleting them depolarizes astrocytes resulting in inhibition of K⁺ and glutamate uptake (Djukic et al., 2007). In addition, several studies report that depolarization of astrocytes can reverse the GlyT1 transporter resulting in elevated extracellular levels of glycine (Roux et al., 2000; Sakata et al., 1997; Shibasaki et al., 2017). We hypothesized that blocking Kir4.1 would depolarize OFC astrocytes, leading to a release of glycine through reversal of GlyT1 and a decrease in spike firing of OFC neurons. We also predicted that this might occlude any further inhibition by ethanol. In these studies, we obtained baseline recordings and those in the presence of the selective Kir4.1 blocker VU0134992 (10 μM), VU plus ethanol (66 mM) followed by a washout. Two-way analysis of these data revealed a significant main effect of treatment (Figure 5A; two-way repeated measures ANOVA: main effect of treatment F_(3,30) = 51.04, p<0.0001; N=11). Post-hoc comparisons showed that spiking was reduced by VU0134992 (Dunnett’s multiple comparison baseline versus VU0134992 q=8.68, p<0.0001). In the presence of VU0134992, ethanol produced an additional decrease in the number of spikes that was significantly different from the VU0134992 baseline (Dunnett’s multiple comparison q=3.29; p=0.0072). Likewise, there were differences in total spike numbers between VU0134992 vs baseline (one-way repeated measures ANOVA, F_(3,30) = 51.04, p<0.0001, Dunnett’s multiple comparisons q=8.68, p<0.0001), and between VU0134992 and ethanol+VU0134992 (Dunnett’s multiple comparison q=3.29, p=0.0072) (Figure 5B). To investigate whether the inhibition of spiking by VU0134992 was due to the release of glycine, we included the glycine receptor antagonist strychnine in the bath solution. Unlike the results shown in Figure 5A, VU0134992 had no effect on current-evoked spiking in the presence of strychnine (Figure 5C; two-way repeated measures ANOVA: main effect of treatment F_(2,18)=0.16, p=0.85; N=10;). Total spike number during VU0134992 application was also not significantly different from strychnine baseline or washout recording (Figure 5D; one-way repeated measures ANOVA F_(2,23)=0.049, p=0.95). These results suggest that blocking Kir4.1 channels inhibits spike firing of lOFC neurons through the release of astrocytic glycine.

Figure 5.

Over-expression of a Ca²⁺ pump in astrocytes together with exposure to a Kir4.1 channel inhibitor suppresses ethanol-induced inhibition of lOFC spiking. (A) In naïve mice, the Kir4.1 blocker VU0134992 (10 μM) significantly decreased OFC neuron spike firing as compared to baseline (two-way repeated measures ANOVA: main effect of treatment F_(3,30) = 51.04, p<0.0001; Dunnett’s multiple comparison baseline versus VU0134992 q=8.68, p<0.0001, N=11, 3 mice). Application of ethanol in the presence of VU0134992 further reduced number of spikes of OFC neurons (Dunnett’s multiple comparison q=3.29; p=0.0072). (B) There were differences in total spike numbers between VU0134992 vs baseline (one-way repeated measures ANOVA, F_(3,30) = 51.04, p<0.0001, Dunnett’s multiple comparisons q=8.68, p<0.0001), and between VU0134992 vs ethanol+VU0134992 (Dunnett’s multiple comparison q=3.29, p=0.0072). (C) VU0134992 failed to induce a decrease in AP spiking in the presence of the glycine receptor blocker strychnine (1 μM), (two-way repeated measures ANOVA: main effect of treatment F_(2,18)=0.16, p=0.85; N=10, 3 mice). (D) Total spike number during VU0134992 application was also not significantly different from baseline (one-way repeated measures ANOVA F_(2,23)=0.049, p=0.95). (E) Baseline firing of OFC neurons in slices from mice injected with the AAV2/5 GfaABC1D hPMCA2w/b virus and treated with VU0134992 for the entire recording duration. In the presence of both AAV2/5 GfaABC1D hPMCA2w/b and VU0134992, ethanol did not inhibit OFC spiking (two-way repeated measures ANOVA: F_(1,13) = 0.042, p=0.84; N=14, 4 mice). (F) There was no difference in total spike number during ethanol+VU0134992 as compared to VU0134992 baseline (one-way repeated measures ANOVA F(2,23)=0.049, p=0.95). Symbols: (*) p<0.05, (****) p<0.0001.

We also tested the effect of VU0134992 on ethanol inhibition of OFC neurons in mice injected with AAV-GfABC1D-hPMCA2w/b. In this experiment, OFC slices were perfused with aCSF containing VU0134992 (10 μM) for the entire recording period. In the presence of the Kir4.1 blocker alone, all cells patched in slices from AAV-GfABC₁D-hPMCA2w/b injected animals showed reduced spike number similar to that obtained when VU0134992 was acutely applied to control slices (compare Figure 5A and Figure 5E). However, under these conditions (hPMCA2w/b plus VU0134992), ethanol had no effect on OFC neuron firing (two-way repeated measures ANOVA: F_(1,13) = 0.042, p=0.84; N=14; Figure 5E). Similarly, in cells from hPMCA2w/b-injected mice, there was no difference in total spike number during ethanol+VU0134992 as compared to VU0134992 baseline (Figure 5D; one-way repeated measures ANOVA F(2,23)=0.049, p=0.95). As Kir4.1 is only expressed in astrocytes, these results suggest that Kir4.1 channels are partially involved in ethanol-induced release of glycine, presumably through the reversal of astrocytic GlyT1. These results also support that proper intracellular astrocytic calcium signaling is important for the action of ethanol on OFC neuron intrinsic excitability.

3.5. Inhibitor of phospholipase C or conventional protein kinase C suppresses ethanol inhibition of lateral OFC neuron excitability

The dopamine D1/D5 receptor is classically thought to be coupled to the Gsα subunit of the G protein, however, D1-like receptors have also been shown to couple to the Gq/PLC pathway (Undie and Friedman, 1990; Undieh, 2010; Medvedev et al, 2013). It was previously reported that, in cortical astrocytes, the dopamine D1/D5 receptor induced reversal of GlyT1 was mediated by a Gq-PLC/PKC pathway (Shibasaki et al., 2017). In addition, application of a PKC activator has been shown to reduce the open state probability of Kir4.1/Kir5.1 channels (Rojas et al., 2007). Therefore, in light of the findings discussed above, it is possible that ethanol inhibition of OFC spike firing involves activation of a D1/D5 receptor/PLC/PKC pathway that leads to inhibition of astrocyte Kir4.1 channels. To test this hypothesis, we examined the ability of ethanol to inhibit OFC firing following treatment of slices with an inhibitor of PLC (U73122), or agents that inhibit broad spectrum subtypes of PKC (chelerythrine) or conventional forms (Gö6976). EtOH (66 mM) alone induced a significant inhibition of AP spiking of OFC neurons as compared to baseline (two-way repeated measures ANOVA: main effect of treatment, F_(2,18) = 12.20, p=0.0004; Dunnett’s multiple comparison baseline versus ethanol q=4.78, p=0.0003, N=10). Total spike number was also significantly reduced by ethanol (Figure 6B, one-way repeated measures ANOVA F_(2,17)=12.20, p=0.0005, Dunnett’s multiple comparison q=4.78, p=0.0004) and was 42.29 ± 11.67% of the control value. In the presence of the PLC inhibitor U73122 (5 μM), ethanol had no significant effect on OFC neuron spiking (Figure 6C, two-way repeated measures ANOVA: main effect of treatment F_(2,24) = 3.6, p=0.07, Dunnett’s multiple comparison U73122 versus U73122 plus ethanol q=0.87, p=0.60; N=13). By itself, U73122 did not alter OFC spiking as compared to the pre-drug baseline (Dunnett’s multiple comparison baseline versus U73122 q=1.57, p=0.22). There were no differences in total spike number among these conditions (Figure 6D, one-way repeated measures ANOVA, F_(2,24) = 3.06, p=0.07). When recordings were performed in the presence of the PKC inhibitor chelerythrine (10 μM) that by itself had no effect (Figure 6E; two-way repeated measures ANOVA, main effect of treatment F_(3,27)=0.09, p=0.96; Dunnett’s multiple comparison baseline versus Chelerythrine q=0.21, p>0.05), ethanol no longer inhibited spike firing (Dunnett’s multiple comparison Chelerythrine versus Chelerythrine plus ethanol q=0.05, p>0.05; N=9). Total spike numbers among these conditions were not significantly different (Figure 6F; one-way repeated measures ANOVA, F_(2,15) = 0.72, p=0.50). Similarly, when the conventional PKC with cPKC inhibitor Gö6976 (1 μM) was present during the recordings, ethanol had no effect on OFC firing (Figure 6G; two-way repeated measures ANOVA: main effect of treatment F_(3,18) = 2.47, Dunnett’s multiple comparison Go6976 versus ethanol q=0.022, p>0.05; N=7; Figure 6G). Gö6976 alone did not alter AP spiking as compared to baseline (q=2.02, p>0.05). Similarly, total spike numbers were not different from one another (one-way repeated measures ANOVA, F_(3,15) = 2.47, p=0.10; Figure 6H). These results suggest the involvement of PLC/cPKC pathway in the effects of ethanol on lOFC neuron firing.

Figure 6.

Inhibition of PLC or cPKC suppresses ethanol-induced inhibition of OFC spiking. (A) Ethanol (66 mM) alone significantly decreased spike firing of OFC neurons as compared to baseline (two-way repeated measures ANOVA: main effect of treatment, F_(2,18) = 12.20, p=0.0004; Dunnett’s multiple comparison baseline versus ethanol q=4.78, p=0.0003, N=10, 6 mice). (B) Total spike number during ethanol treatment was significantly different from baseline (one-way repeated measures ANOVA F_(2,17)=12.20, p=0.0005, Dunnett’s multiple comparison q=4.78, p=0.0004). (C) In the presence of the PLC inhibitor U73122 (5 μM), no decrease of OFC firing by ethanol was observed (two-way repeated measures ANOVA: main effect of treatment F_(2,24) = 3.6, p=0.07, Dunnett’s multiple comparison U73122 versus U73122 plus ethanol q=0.87, p=0.60; N=13). U73122 alone did not alter OFC spiking as compared to the pre-drug baseline (Dunnett’s multiple comparison baseline versus U73122 q=1.57, p=0.22). (D) There were no differences in total spike number among these conditions (one-way repeated measures ANOVA, F_(2,24) = 3.06, p=0.07). (E) In the presence of the PKC inhibitor chelerythrine (10 μM) that by itself did not affect neuron firing (two-way repeated measures ANOVA, main effect of treatment F_(3,27)=0.09, p=0.96; Dunnett’s multiple comparison baseline versus Chelerythrine q=0.21, p>0.05), ethanol did not decrease AP spiking of OFC neurons (Dunnett’s multiple comparison Chelerythrine versus Chelerythrine plus ethanol q=0.05, p>0.05; N=9, 3 mice). (F) Likewise, total spike numbers at baseline and during chelerythrine and ethanol plus chelerythrine were not different (one-way repeated measures ANOVA, F_(2,15) = 0.72, p=0.50). (G) In the presence of the cPKC inhibitor Gö6976 (1 μM), ethanol failed to suppress ethanol inhibition of OFC firing (two-way repeated measures ANOVA: main effect of treatment F_(3,18) = 2.47, Dunnett’s multiple comparison Go6976 versus ethanol q=0.022, p>0.05; N=7, 3 mice). Gö6976 alone did not alter AP spiking as compared to baseline (q=2.02, p>0.05). (H) Total spike numbers were not different from one another (one-way repeated measures ANOVA, F_(3,15) = 2.47, p=0.10). Symbol: (***) p<0.001.

3.6. Ethanol, Kir4.1 blockers or a D1/D5 receptor agonist depolarize OFC astrocytes.

Inhibition of Kir4.1 channels has been shown to depolarize astrocytes (Djukic et al., 2007) that may lead to a reversal of the GlyT1 transporter and increased extracellular levels of glycine (Roux et al., 2000; Sakata et al., 1997; Shibasaki et al., 2017). To test how various modulators used in this study affect astrocyte membrane potential, we treated OFC slices with sulforhodamine 101 (SR101) and used sharp electrodes filled with 1 M KCl internal solution to monitor changes in membrane potential of OFC astrocytes (see methods). Following electrode penetration, we recorded from astrocytes until a stable baseline was achieved (~15–20 min). Then, the bath solution was switched to one containing either barium that blocks a variety of K⁺ channels, the selective Kir4.1 blocker VU0134992, the D1R-like agonist SKF81297 or ethanol for 20 min followed by a washout for another 20 min. In control recordings in regular aCSF, the RMP of astrocyte at 15 min of recording was −77.95 ± 1.72 mV, and this value was stable over a 55 min recording period (Figure 7A; one-way repeated measures ANOVA, F_(8,56) = 0.82, p=0.59; N=8). In separate recordings, application of barium (100 μM) significantly depolarized the astrocyte membrane potential from −78.55 ± 4.33 mV at baseline to −62.43 ± 3.51 mV (Figure 7A; one-way repeated measures ANOVA F_(8,56)=12.80, p<0.0001; Dunnett’s multiple comparisons 15 min value versus 20 min, p=0.018; 15 min versus 25, 30, 35 min, p<0.0001). This membrane potential was not reversed after 20 min of washout (Dunnett’s multiple comparisons; 15 min versus 40, 45, 50, 55 min, p<0.0001). Prior to application of VU0134992 (10 μM), the RMP of astrocytes at 20 min was −83.24 ± 4.67 mV (Figure 7B). VU0134992 (10 μM) significantly depolarized the astrocyte membrane potential to −66.98 ± 3.97 mV (one-way repeated measures ANOVA, F_(8,32) = 6.73, p<0.0001; N=5; Dunnett’s multiple comparisons 20 min versus 30 min, p=0.014; 20 min versus 35, p=0.0002). The membrane potential did not recover after washout of VU0134992 (Dunnett’s multiple comparisons 20 min versus 40, 45 min, p<0.0001; 20 min versus 50 min p=0.0004, 20 min versus 55, 60 min, p=0.002). In the experiment with the D1 agonist SKF81297 (Figure 7C), the astrocyte RMP at baseline (15 min) was −79.39 ± 4.74 mV. During exposure to SKF81297 (10 μM), the membrane potential was significantly depolarized to −65.23 ± 5.58 mV and did not recover following washout (one-way repeated measures ANOVA, F_(8,56) = 17.83, p<0.0001; Dunnett’s multiple comparisons 15 min versus 25–55 min, p<0.0001). In the ethanol recordings, the RMP of lOFC astrocytes was −75.87 ± 1.94 mV at the end of the baseline period (Figure 7C). Application of ethanol (66 mM) significantly depolarized the membrane potential to −66.68 ± 2.69 mV (one-way repeated measures ANOVA, F_(8,40) = 10.98, p<0.0001; N=6) and this did not recover after washout (Dunnett’s multiple comparisons 15 min value versus 20 min, p=0.02; 15 min versus 25–55, p<0.0001). To test whether the ethanol-induced depolarization was dopamine dependent, the astrocyte RMP was monitored under baseline conditions and during application of the D1R antagonist SCH23390 and then ethanol in the presence of SCH23390. As shown in Figure 7D, application of the D1 antagonist SCH23390 (10 μM) slightly depolarized the astrocyte membrane potential from −69.73 ± 3.73 mV at baseline (15 min) to −67 ± 3.52 mV (35 min) (one-way repeated measures ANOVA, F_(12,72) = 4.311, p=0.028; N=7, 5 mice). When added in the presence of SCH23390, ethanol (66 mM) had no additional effect on the astrocyte membrane potential (Dunnett’s multiple comparisons 35 min value versus 40–55 min, p>0.05). Finally, to examine whether the effects of ethanol on astrocyte membrane potential and current-evoked spiking described above are found in other cortical areas, we performed recordings in deep-layers of the prelimbic medial prefrontal cortex (mPFC). Under baseline condition (15 min), the RMP of mPFC astrocytes was −81.4 ± 3.26 mV (Figure 7E). However, unlike lOFC astrocytes, ethanol (66 mM) had no effect on the membrane potential of mPFC astrocytes (−80.26 ± 3.46 mV at 35 min; one-way repeated measures ANOVA, F_(8,32) = 3.416, Dunnett’s multiple comparisons 15 min value versus 20–35 min, p>0.05; N=5). Similarly, while ethanol consistently inhibited current-evoked spiking of lOFC neurons, it had no effect on spike firing of mPFC neurons (Figure 7F; two-way repeated measures ANOVA: main effect of treatment, F_(2,16) = 1.822, p=0.19; N=9).

Figure 7.

Ethanol, Kir4.1 blockers or a D1/D5 receptor agonist depolarize OFC astrocytes. (A) In regular aCSF (control), the RMP of astrocyte was stable over a 55 min recording period (one-way repeated measures ANOVA, F_(8,56) = 0.82, p=0.59; N=8, 4 mice). Barium (100 μM) significantly depolarized astrocyte membrane potential as compared to baseline and this was not reversed after 20 min washout (−64.19 ± 3.73 mV; one-way repeated measures ANOVA F_(8,56)=12.80, p<0.0001; Dunnett’s multiple comparisons 15 min value versus 20 min, p=0.018; 15 min versus 25–55min, p<0.0001; N=8, 5 mice). (B) VU0134992 significantly depolarized astrocyte membrane potential and this did not recover during washout (one-way repeated measures ANOVA, F_(8,32) = 6.73, p<0.0001; Dunnett’s multiple comparisons 20 min versus 30 min, p=0.014; 20 min versus 35, p=0.0002, 20 min versus 40, 45 min, p<0.0001; 20 min versus 50 min p=0.0004, 20 min versus 55, 60 min, p=0.002; N=5, 3 mice. (C) As compared to baseline, SKF81297 (10 μM) significantly depolarized astrocyte membrane potential and this did not recover during washout (one-way repeated measures ANOVA, F_(8,56) = 17.83, p<0.0001; Dunnett’s multiple comparisons 15 min versus 25–55 min, p<0.0001; N=8, 3 mice). In the ethanol experiment, there was a significant depolarization of OFC astrocyte membrane potential by ethanol (66 mM) as compared to baseline RMP (one-way repeated measures ANOVA, F(8,40) = 10.98, p<0.0001; N=6) and this did not recover after washout (Dunnett’s multiple comparisons 15 min value versus 20 min, p=0.02; 15 min versus 25–55 min, p<0.0001). (D) In the presence of the D1 antagonist SCH23390 (10 μM) that by itself slightly depolarized the RMP of astrocytes (one-way repeated measures ANOVA, F_(12,72) = 4.31, p=0.028; N=7), ethanol (66 mM) had no effect on the astrocyte membrane potential (Dunnett’s multiple comparisons 35 min value versus 40–55 min, p>0.05). (E) In mPFC astrocyte recordings, ethanol (66 mM) did not alter the membrane potential of mPFC astrocytes as compared to 15-min baseline (one-way repeated measures ANOVA, F_(8,32) = 3.416, Dunnett’s multiple comparisons 15 min value versus 20–35 min, p>0.05; N=5). (F) Effect of ethanol (66 mM) on AP spiking (mean ± SEM) induced by a series of current injections (40–220 pA) in prelimbic mPFC neurons. In the presence of ethanol, there was no change in spike firing of prelimbic mPFC neurons as compared to control baseline (two-way repeated measures ANOVA: main effect of ethanol, F_(2,16) = 1.822, p=0.19; N=9).

4. Discussion

We reported previously that ethanol, applied acutely, reduces the intrinsic excitability of lOFC neurons via the activation of glycine receptors (Badanich et al., 2013). Following repeated cycles of CIE exposure, firing is enhanced and bath application of ethanol no longer inhibits spike firing or generates glycine-mediated tonic currents (Nimitvilai et al., 2016). In addition, like ethanol, dopamine or a D1/D5 agonist increases the holding current of lOFC neurons in naïve mice (Nimitvilai et al., 2017). In this study, we examined possible mechanisms linking ethanol and dopamine to the glycine-dependent inhibition of lOFC neuron spiking. Our data demonstrate that ethanol inhibition of lOFC neuron excitability is dependent on activation of the D1/D5 receptor and is blocked by antagonists of norepinephrine and glycine transporters. Manipulating astrocyte function by blunting astrocytic Ca²⁺ signaling or blocking astrocyte-selective Kir4.1 channels diminished but did not completely eliminate ethanol-induced inhibition of OFC spike firing. However, combining Ca²⁺ extrusion with Kir4.1 inhibition totally prevented ethanol inhibition of firing as did an inhibitor of Ca²⁺-dependent conventional PKC. Finally, bath application of either barium, a Kir4.1 blocker, D1 agonist or ethanol all depolarized OFC astrocytes while a D1 antagonist blocked the ethanol effect. While other as yet to be identified mechanisms may be involved, these results suggest that acute ethanol inhibits OFC neuron excitability via dopamine-induced activation of D1/D5 receptors that induces a cPKC-dependent inhibition of Kir4.1 channels, depolarization of astrocytes and reversal of GlyT1-mediated transport (Figure 8). Importantly, the effects of ethanol on lOFC astrocytes and neurons were not observed in the prelimbic mPFC suggesting that these regions likely experience different outcomes during voluntary consumption of ethanol.

Figure 8.

Schematic showing proposed model of action of ethanol. Ethanol acts on the norepinephrine transporter to release dopamine (1). Dopamine then activates D1/D5-coupled Gq receptors on astrocytes (2) that subsequently stimulates a PLC/cPKC pathway (3). PLC/cPKC pathway triggers the reversal of astrocytic GlyT1 directly (solid arrow, 4) and indirectly via membrane depolarization that occurs following inhibition of Kir4.1 channels (dashed arrow, 5). Glycine released during reverse GlyT1 transport binds and stimulates inhibitory strychnine-sensitive glycine receptors (6), resulting in a decrease in the intrinsic excitability of lOFC neurons. Symbols: (*) p<0.05, (**) p<0.01, (***) p<0.001, (****), p<0.0001.

4.1. Ethanol and Monoamine Transporters

Results from the present study show that ethanol inhibition of OFC firing was prevented by inhibitors of monoamine transporters. Previous studies show that acute ethanol potentiated dopamine uptake in dopamine transporter (DAT) transfected oocytes, and this was associated with increased DAT expression at the cell surface (Mayfield et al., 2001). Acute exposure to ethanol also enhances serotonin transporter (SERT) activity in rat cortical, hippocampal and brainstem synaptosomes (Alexi and Azmitia, 1991). However, the norepinephrine transporter (NET) has been reported to be inhibited by acute ethanol in rat cerebellum and cortical slices (Israel et al., 1973; Lin et al., 1993). In the frontal cortex, NETs and SERTs are highly expressed, while that of DAT is sparse (Freed et al., 1995; Morón et al., 2002; Sesack et al., 1998). Although SERT’s weak affinity for dopamine likely prevents it from taking up that neurotransmitter at physiological levels (Hoffman et al., 1991), NET has greater affinity for dopamine than DAT itself and plays a critical role in regulating extracellular levels of dopamine in the frontal cortex (Morón et al., 2002; Mundorf et al., 2001; Yamamoto et al., 1998). This is supported by studies showing that selective blockers of NET, but not DAT, inhibit dopamine uptake in the PFC (Carboni et al., 1990; Morón et al., 2002; Tanda et al., 1994). In addition, dopamine uptake by frontal cortex synaptosomes is normal in DAT-knockout mice, while it is significantly decreased in NET-deficient mice (Morón et al., 2002). Monoamine transporters, including NET, can operate in reverse under certain conditions leading to release of dopamine (Sonneborn and Greene, 2019). Reverse transport can be driven by increases in intracellular Na⁺ concentrations (Aubrey et al., 2005) and acute ethanol has been reported to elevate intracellular Na⁺ by inhibiting the Na⁺/K⁺ ATPase pump (Israel et al., 1973; Ledig et al., 1985). While some studies report that relatively high concentrations (125 mM to 2 M) of ethanol are required for this effect (Rothman et al., 1996; Swann, 1990; Syapin et al., 1985), others show that suppression of Na+/K+ ATPase activity occurs at low to moderate concentrations (40–100 mM) (Botta et al., 2010; Israel et al., 1973; Ledig et al., 1985). In this study, the selective NET inhibitor nisoxetine blunted ethanol inhibition of current-evoked OFC neuron spiking, suggesting that ethanol may induce reversal of NET causing an increase in extracellular levels of dopamine. Future studies could examine whether in the OFC, acute ethanol causes a NET-dependent release of DA via inhibition of Na⁺/K⁺ ATPase.

4.2. Dopamine and GlyT1

In the present study, we show that dopamine enhanced a glycine-mediated tonic current; that the ethanol inhibition of OFC excitability was suppressed by a non-transportable GlyT1 inhibitor; and that ethanol and a D1 agonist both depolarized the lOFC astrocyte membrane potential. Since GlyT1 is highly expressed in glial cells (Adams et al., 1995), these results suggest that ethanol and dopamine may share a common pathway that stimulates the release of glycine, possibly through a reversal of astrocytic GlyT1. Under physiological conditions, astrocytic transporters take up excessive neurotransmitters from synaptic sites. However, when intracellular concentrations of particular neurotransmitters are increased, astrocytic transporters can release gliotransmitters and this is thought to be independent of astrocytic Ca²⁺ concentration (Eulenburg et al., 2005; Huang et al., 2004; Roux et al., 2000; Sakata et al., 1997). In cortical astrocytes, dopamine has been shown to cause a functional reversal of GlyT1 via activation of a D5 receptor/PLC pathway, resulting in the release of glycine from astrocytes (Shibasaki et al., 2017). These findings are consistent with findings from the present study and suggest that manipulating astrocyte function should interfere with ethanol inhibition of OFC spike firing. Although early immunohistochemical studies indicated that GlyT1 expression is restricted to glia (Adams et al., 1995; Zafra et al., 1995), GlyT1 has since been reported on some vGluT1 positive presynaptic vesicles and glutamatergic synapses where it may assemble with NMDA receptors (Cubelos et al., 2005; Cubelos et al., 2014). Thus, although astrocytes appear to be the most obvious source of the glycine implicated by the results of the present study, we cannot rule out a potential role of neuronal GlyT1 or glycine in the effects of ethanol on lOFC neuronal excitability.

4.3. Role of Astrocytes in ethanol Inhibition of lOFC Neuron Firing

Disrupting Ca²⁺ signaling in astrocytes has been shown to affect neuronal activity and impaired astrocyte function is implicated in various neurological disorders (Kuchibhotla et al., 2009; Sofroniew and Vinters, 2010). In a recent study, over-expression of a human plasma membrane calcium pump (hPMCA2w/b) in mouse striatal astrocytes significantly decreased spontaneous and evoked astrocytic Ca²⁺ signals and reduced current-evoked firing of striatal medium spiny neurons (Yu et al., 2018). In the present study, AAV-mediated expression of the same hPMCA2w/b construct in lOFC astrocytes reduced but did not completely eliminate ethanol inhibition of OFC neuron firing. These results suggest that either an additional mechanism is involved in ethanol inhibition of firing or that a more complete block of astrocyte function is required to completely eliminate ethanol’s effect on OFC neuron excitability.

One of the main functions of astrocytes is to regulate the extracellular concentration of K⁺, through a process called K⁺ spatial buffering. During periods of neuronal activity, astrocytes take up excess extracellular K⁺, distribute it to other astrocytes via gap junctions and then extrude the excess ions at sites of lower K⁺ concentrations (Kuffler and Nicholls, 1966). Kir4.1 channels are the main regulator of K⁺ in glial cells, are not expressed in neurons (Brasko et al., 2017; Butt et al., 2006; Higashi et al., 2001) and are implicated in alcohol addiction. For example, expression of the gene encoding Kir4.1, Kcnj10, is altered in animals following chronic alcohol drinking and in brain tissue from humans with alcohol use disorder (Buck and Finn, 2001; Buck et al., 1997; Lewohl et al., 2000; Tarantino et al., 1998). In addition, downregulation of brain kcnj10 transcripts has been reported in ethanol-preferring C57BI/6J mice as compared with ethanol-avoiding BALB/cJ mice (Zou et al., 2009). Kir channels are regulated by a number of neurotransmitters through specific intracellular signaling systems (Kim et al., 1995; Lacey et al., 1987, 1988; Mao et al., 2004) including protein kinase C. Several studies have reported that the Kir4.1 subunit contains consensus sequences encoding potential PKC phosphorylation sites and that activation of PKC inhibits Kir4.1 function (Rojas et al., 2007; Zaika et al., 2013). In this study, a selective Kir4.1 blocker depolarized the astrocyte membrane potential, similar to that reported previously (Djukic et al., 2007), and decreased current-evoked firing of lOFC neurons. Activation of D1-like receptors has been reported to stimulate a Gq/PLC/PKC pathway that subsequently induces glycine release via functional reversal of cortical astrocyte GlyT1 (Shibasaki et al., 2017). In the current study, we found that a PLC inhibitor or a blocker of Ca²⁺-dependent cPKC eliminated the inhibitory effects of ethanol on lOFC firing and like the Kir4.1 blocker, ethanol or the D1 agonist depolarized the membrane potential of OFC astrocytes, thus supporting a role for astrocytic PLC/PKC signaling in these effects. Nonetheless, since these PLC/cPKC blockers are not selective for astrocytes, we cannot rule out the possibility that their effect was mediated via inhibition of neuronal or non-astrocytic forms of PLC/PKC. In the presence of the Kir4.1 blocker, ethanol produced an additional small but significant inhibition of spiking similar to those from hPMCA2w/b expressing mice while blocking Kir4.1 channels in slices from hPMCA2w/b expressing mice totally prevented ethanol inhibition of OFC firing. As expression of hPMCA2w/b was previously reported not to affect levels of Kir4.1 channels in astrocytes (Yu et al., 2018), it is likely that at least two mechanisms, altered astrocytic Ca²⁺ signaling and Kir4.1 inhibition-induced depolarization of astrocytes, are involved in ethanol-induced release of glycine and and subsequent inhibition of OFC firing.

4.4. Conclusions

Dysfunction of astrocytes in the OFC has been reported in ethanol-dependent rats and human alcoholics (Miguel-Hidalgo et al., 2006; Miguel-Hidalgo et al., 2002; Miguel-Hidalgo et al., 2014). We reported previously that application of the transportable GlyT1 inhibitor sarcosine decreases AP spiking of OFC neurons, and this inhibition is lost after CIE exposure (Nimitvilai et al., 2016), suggesting that chronic ethanol exposure may disrupt GlyT1 function or expression. In addition, CIE treatment also blunts the effects of acute ethanol on AP spiking and tonic glycine currents (Nimitvilai et al., 2016). Together with the results in this study, it is possible that CIE exposure impairs astrocytic release of glycine that contributes to the altered excitability of OFC neurons observed in CIE exposed animals. Finding ways to restore astrocyte function might lead to novel interventions for treatment of OFC-dependent behavioral deficits in alcohol-dependent individuals.

Highlights.

• Ethanol inhibits lOFC neuron spiking and depolarizes the astrocyte membrane potential

• Inhibition is blocked by antagonists of monoamine and glycine transporters

• Ethanol inhibition of lOFC neuron spiking is D1/D5 receptor dependent and involves a PLC/cPKC pathway that may target astrocytic selective Kir4.1 potassium channels

• Ethanol-induced depolarization of astrocytes may reverse astrocytic GlyT1 transporters leading to efflux of glycine and activation of strychnine-sensitive GlyRs on lOFC neurons