Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Ethanol alters neural activity by a number of mechanisms, including a direct action on neurotransmitter-gated ion channels (Crews et al., 1997). In vivo, ethanol inhibits the action of NMDA (Simson et al., 1991, 1993; Criswell et al., 1993) and enhances the inhibitory action of GABA (Givens and Breese, 1990a,b; Lin et al., 1993; Criswell et al., 1995). However, not all brain areas are affected equally by ethanol in vivo. Intoxicating doses of ethanol depress neural activity in the medial septum, without an effect on the nearby lateral septum (Givens and Breese, 1990a,b). Furthermore, ethanol blocks the changes in neuronal activity produced by iontophoretic application of NMDA and enhances the changes produced by iontophoretic application of GABA onto neurons in the medial septum, substantia nigra, and inferior colliculus but does not have this effect on neurons from the lateral septum (Simson et al., 1991, 1993; Yang et al., 1996). Ethanol enhances the effect of GABA on GABA_A receptors on cerebellar Purkinje neurons only in the presence of a -adrenergic agonist (Lin et al., 1993; Freund and Palmer, 1997) or a GABA_B agonist (Yang et al., 2000). Thus, there is a brain regional specificity for the effect of ethanol on modulation of GABA and NMDA function in vivo.

Although several in vitro studies have demonstrated inhibition of NMDA-gated currents by ethanol (Lovinger, 1995; Grover et al., 1998; Ming et al., 2001), the brain sites investigated did not include the lateral septum to determine whether ethanol would be without effect at this site. In vitro studies of the effect of ethanol on GABA-gated currents have been equivocal. Although some studies found an enhancement of GABA-gated currents by ethanol (Wafford et al., 1991; Reynolds et al., 1992), others did not find a direct effect of ethanol on GABA-gated currents (White et al., 1990; Sigel et al., 1993; Criswell et al., 1999; Ming et al., 2001) or found enhancement of GABA-gated currents by ethanol in some cases but not others (Weiner et al., 1994; Wan et al., 1996; Sapp and Yeh, 1998). The purpose of the present investigation was to examine the sensitivity to ethanol of NMDA and GABA_A receptors in neurons acutely dissociated from several brain regions to determine whether different postsynaptic receptor sensitivity to ethanol across brain regions might explain the presence or absence of an effect of ethanol on neural activity observed in vivo. Because endogenous neuroactive steroids are present in vivo but not in vitro, the effect of alphaxalone on GABA-gated currents was tested in vitro to determine whether endogenous neurosteroids might contribute to the regionally specific effects of ethanol observed in vivo.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Drugs. Chemicals for producing internal and external recording solutions as well as NMDA and GABA were obtained from Sigma-Aldrich (St. Louis MO). Ifenprodil bitartrate and alphaxalone were obtained from Sigma/RBI (Natick, MA), whereas ethanol (95%) was obtained from Aaper (Shelby, KY).

Preparation of Acutely Dissociated Neurons. Fifteen- to twenty-one-day-old rats were anesthetized with urethane (1.5 g/kg; Sigma-Aldrich) and their brains removed and placed in ice-cold normal saline. Brain areas (substantia nigra, lateral septal nucleus, thalamus, hippocampus, or cerebellum) were dissected out, chopped into approximately 0.5-mm pieces, immersed in standard external solution (SES) for patch recording, and bubbled for 1 h with 100% O₂. The tissue was then transferred into a solution of SES with 3 mg/ml Protease XXIII for 24 to 28 min. The tissue was then rinsed and stored in oxygenated SES until recording. At the time of recording, a piece of tissue was triturated by a fire-polished Pasteur pipette and the dissociated cells were allowed to settle into a recording chamber. Recordings were made within 1 h of plating the cells.

Transfection of CHO Cells. CHO-K1 cells were cultured on poly-L-lysine (0.04 mg/ml)-coated, 12-mm glass coverslips in a standard 12-well plate at 37°C, 5% CO₂ and allowed to reach 50 to 80% confluence. Culture media consisted of Ham's F-12 + 10% fetal bovine serum, 100 U/ml penicillin, and 0.1 mg/ml streptomycin (Invitrogen, Carlsbad, CA). Cells were rinsed once with phosphate-buffered saline and then transfected for 5 to 6 h with 750 µl/well Opti-MEM1 (Invitrogen) containing a mixture of 3 µl/well LipofectAMINE reagent (Invitrogen) and a total of 1.3 µg/well plasmid DNA. The plasmid DNA mixture consisted of 0.1 µg/well pEGFPN1 (green fluorescent protein), prC/CMV-NR1-1a, or prC/CMV-NR1-1b (0.6 µg/well), and NR2 subunits (prC/CMV-NR2a or prC/CMV-NR2b) alone or in a 1:1 ratio (0.6 µg/well when alone, 0.3 µg/well when together). NMDA receptor plasmids were a generous gift from Dr. David Lovinger (Vanderbilt University, Nashville, TN). After 5 to 6-h incubation, the transfection mixture was removed and the cells were fed with 1 ml/well fresh media (Ham's F-12 + 10% fetal bovine serum, 100 U/ml penicillin, and 0.1 mg/ml streptomycin). MK-801 (100 µM; Sigma-Aldrich) or DL-2-amino-5-phosphonovaleric acid (1 mM; Sigma-Aldrich) was added to the media to prevent glutamate toxicity. Transfected cells were used for electrophysiological recording for 1 to 3 days after transfection. Positively transfected cells were identified by green fluorescent protein fluorescence.

Culture of Ltk Cells. Ltk cells (generously donated by Dr Paul Whiting, Merck Sharp and Dohme, Terlings Park Harlow, Esex, UK) were cultured in Dulbecco's modified Eagle's medium (Invitrogen) on poly-L-lysine (0.04 mg/ml)-coated, 12-mm glass coverslips in a standard 12-well plate at 37°C, 5% CO₂ and allowed to reach 60 to 80% confluence. One to 3 days before electrophysiological recording, the cells were induced to trigger expression of GABA_A receptors by adding 1 µM dexamethasone.

Electrophysiological Recording. Electrophysiological studies were performed under voltage clamp in the whole-cell configuration using an Axopatch-1D amplifier (Axon Instruments, Inc., Foster City, CA). Recording pipettes were fabricated from N 51A capillary glass (Drummond Scientific, Bromall, PA). The internal solution used for measuring ion currents induced by NMDA or GABA included 150 mM KCl, 15 mM HEPES, 2 mM K-ATP, 5 mM EGTA, 15 mM phosphocreatine, and 50 U/ml creatine phosphokinase. Inclusion of the last two items regenerates ATP and GTP and decreases rundown of ion channel-mediated currents (Forscher and Oxford, 1985). This solution was adjusted to pH 7.4 and had an osmolality of 310 (adjusted with sucrose). In some cases, the KCl was replaced by CsCl and was pH adjusted with CsOH. Seals were formed on the neurons with electrodes having a tip resistance of 2 to 4 M. Data were displayed on an oscilloscope, digitized at 50 ms/sample, and stored on a personal computer. Recordings were performed at room temperature in a bath where the neurons were superfused at 0.5 to 1 ml/min with Mg²⁺-free SES (145 mM NaCl, 5 mM KCl, 10 mM HEPES, 2 mM CaCl₂, and 10 mM glucose; pH 7.4, 340 mOsM/kg). NMDA was prepared daily from stock solution (50 mM) and diluted in a batch so that all conditions used the same concentration of NMDA. Ethanol and ifenprodil or alphaxalone were then added to the stock NMDA solution such that they comprised less than 1% of the total volume. Because the diffusion characteristics of solutions are altered by ethanol, solutions were thoroughly stirred after chemicals were added to ensure proper mixing. Solutions were applied by a six-channel rapid translation perfusion system with SES or the various drug combinations present in separate fused silica (180 µm i.d.) tubes positioned 50 to 100 µm from the recorded neuron so as to flood the area surrounding the neuron with a specific drug combination. Position of the tubes could be changed in ~10 ms, thus allowing rapid solution switching and drug application.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effect of Ethanol on NMDA-Gated Currents from Neurons of the Lateral Septum, Hippocampus, Thalamus, and Substantia Nigra. NMDA gated an inward current in neurons acutely isolated from the lateral septum, hippocampus, thalamus, and substantia nigra (Fig. 1). Ethanol produced a concentration-dependent inhibition of the NMDA-gated current in each of these brain regions (Fig. 1). There was a significant effect on NMDA-gated currents of both ethanol concentration and brain region (P < 0.01). Post hoc Tukey's HSD tests showed that the lateral septum was more sensitive to inhibition of NMDA-gated currents by ethanol than the thalamus (P < 0.05). No other regional differences in sensitivity of NMDA-gated currents to ethanol were significant (P > 0.1). An example of the effect of ethanol on NMDA-gated currents is shown in Fig. 2.

View larger version (28K): [in this window] [in a new window]   Fig. 1.   Effect of ethanol on NMDA-gated currents from several brain sites. Ethanol was applied to neurons acutely dissociated from the substantia nigra (sn), lateral septum (ls), hippocampus (hip), and thalamus (thal). Similar concentration-related inhibitions of NMDA-gated currents were recorded from all sites tested. Each column represents data from 11 to 20 neurons.

View larger version (17K): [in this window] [in a new window]   Fig. 2.   Representative examples of NMDA-gated currents from a lateral septal neuron (A) and a thalamic neuron (B). NMDA gated an inward current in both neurons that was blocked in a concentration-dependent manner by ethanol. Ethanol (25 mM, 25) was without a statistically reliable effect, whereas 50 mM ethanol (50) and 100 mM (100) ethanol reliably inhibited the current.

Effect of Ethanol on NMDA-Gated Currents from CHO Cells Expressing NMDA Receptor Subunits. Because a correlation between the presence of the NR2B receptor and ethanol sensitivity had been suggested (Lovinger, 1995; Yang et al., 1996), the effect of ethanol on NMDA-gated currents from CHO cells transfected with NR1 andNR2A and/or NR2B subunits was examined. Figure 3 shows that all of the tested subunit combinations showed similar effects of ethanol on NMDA-gated currents.

View larger version (27K): [in this window] [in a new window]   Fig. 3.   Effect of NMDA on CHO cells transfected with various combinations of NMDA subunits. NMDA (100 µM; cont) or NMDA with 25, 50, or 100 mM ethanol was applied to CHO cells expressing the NR1-1a (1a) or the Nr1-1b (1b) subunit along with the NR2A (2a), the NR2B (2b), or the NR2A and NR2B together (2a2b). Data for each column represent 6 to 20 neurons.

Effect of Ethanol on Ifenprodil-Sensitive and Ifenprodil-Insensitive Currents from the Lateral Septal Nucleus and the Substantia Nigra. Because previous work has shown differences in the effect of ethanol on ion channels expressed in neurons compared with non-neuronal cell lines (Sapp and Yeh, 1998), we tested the effect of ethanol in the presence and absence of ifenprodil to determine whether blocking the NR2B-containing receptors with ifenprodil would alter the effect of ethanol on neurons in the substantia nigra and lateral septum. To verify that ethanol acted on both NR2B-containing receptors and on receptors that did not contain the NR2B subunit in neurons, it was necessary to determine a concentration of ifenprodil selective for the NR2B subunit under our experimental conditions. Therefore, the effect of ifenprodil on NMDA-gated currents was tested in CHO cells transfected with cDNAs for the NR1-1a subunit and either the NR2A or NR2B subunit. This established that 10 µM ifenprodil blocked ~80 to 90% of the current in cells expressing NR2B subunits, but only ~5 to 10% of the current in cells expressing NR2A subunits (Fig. 4).

View larger version (14K): [in this window] [in a new window]   Fig. 4.   Effect of ifenprodil on NMDA-gated currents from NR2A and NR2B-subunit containing cells. Ifenprodil (10 µM) produced a significantly greater inhibition of NMDA-gated currents from CHO cells expressing the NR1-1a subunit with the NR2B (1a2B), compared with the NR2A (1a2A) subunit (P < 0.01; Tukey's HSD). Similar results were obtained from CHO cells expressing the NR1-1b subunit with either the NR2A (1b2A) or the NR2B (1b2B) subunit (P < 0.01; Tukey's HSD). Each column represents data from 13 to 21 cells.

View larger version (21K): [in this window] [in a new window]   Fig. 5.   Effect of ethanol on NMDA-gated currents in the presence or absence of ifenprodil. In the substantia nigra (sn), ethanol (100 mM) inhibited NMDA-gated currents to a similar extent in the presence (ifen) or absence (cont) of ifenprodil. However, in the lateral septum (ls), the effect of ethanol was significantly attenuated by ifenprodil. Each column represents data from 14 to 16 neurons. *, P < 0.05 compared with control.

View larger version (19K): [in this window] [in a new window]   Fig. 6.   Effect of ethanol on NMDA-gated currents in the presence and absence of ifenprodil from CHO cells transfected with various combinations of NMDA receptor subunits. Ethanol (100 mM) inhibited NMDA-gated currents in CHO cells expressing either an NR1-1a (1a) or an NR1-1b (1b) subunit in combination with an NR2-A (2a), NR2-B (2b), or a combination of NR2-A and NR2-B subunits (2a2b). There were no significant differences between the overall effect on all NMDA-gated channels (e100) and those remaining in the presence of 10 µM ifenprodil (e+ifin). The relatively large error bars for the effect of ethanol on NR2-B-containing cells in the presence of ifenprodil are due to an 85 to 90% inhibition of the current by the ifenprodil.

View larger version (24K): [in this window] [in a new window]   Fig. 7.   Examples of inhibition of NMDA-gated currents by ifenprodil and ethanol from CHO cells expressing different combinations of NMDA subunits. Ifenprodil (10 µM) and ethanol (100 mM) were applied to CHO cells expressing NR1-1b and NR2-A subunits (A), NR1-1b and both NR2-A and NR2-B subunits (B), or NR1-1b and NR2-B subunits (C). In each case, NMDA gates a current (N) that is inhibited by ifenprodil (I) to differing degrees, depending on the relative abundance of the NR2-B subunit. The NMDA-gated current is further inhibited when ethanol (E) is added to the NMDA + ifenprodil mixture. When the ethanol is removed and the cell is returned to only NMDA (N) the ethanol effect disappears, but the blockade by ifenprodil remains for several seconds. When the cell is returned to SES (O) the current returns to zero. For comparison, D shows the effect of ethanol on an NMDA-gated current recorded from a CHO cell expressing NR1-1b and NR2-2B subunits in the absence of ifenprodil.

Effect of Ethanol on GABA-Gated Currents. GABA opens a chloride channel and neurons from differing brain areas show different sensitivity to GABA. To ensure that the effect of ethanol was being tested at equivalent concentrations of GABA, we first examined the concentration-response curve for the effect of GABA on neurons from differing brain areas and from CHO cells expressing differing GABA_A receptor subtypes. Figure 8 shows that responses fall into two categories. The cerebellar Purkinje cells and the two cell lines were more sensitive to GABA than were the neurons from substantia nigra or lateral septum. Therefore, differing concentrations of GABA were used such that the control response to GABA did not show more than 20% desensitization during a 4-s application. This resulted in using 1.25 µM GABA for the cerebellar Purkinje neurons and Ltk cells and 5 µM GABA for all other cell types. These concentrations represented approximately an ED₁₅.

View larger version (22K): [in this window] [in a new window]   Fig. 8.   Concentration-response curves for the effect of GABA on neurons acutely isolated from several brain sites. GABA was applied to neurons acutely isolated from the lateral septum (LSN), substantia nigra (SNR), and cerebellar Purkinje neurons (PKZ), and from Ltk cells expressing subunits for the 1, 2 and 2L (2L) or 2S (2S) subunits. All values were normalized to the response at 1 mM.

View larger version (29K): [in this window] [in a new window]   Fig. 9.   Effect of ethanol on GABA-gated currents. GABA was applied at an EC15 concentration to neurons isolated from the substantia nigra (snr), lateral septum (ls), hippocampus (hip), thalamus (thal), and cerebellar Purkinje neurons (pkj). Responses to GABA alone were compared with responses to GABA in the presence of varying concentrations of ethanol (25-300 mM). Control (cont) currents represent the percentage of change between two repeated presentations of GABA alone. Ethanol did not significantly alter the response to GABA at any site tested (P > 0.10). Each column represents data from 9 to 17 neurons.

View larger version (34K): [in this window] [in a new window]   Fig. 10.   Effect of ethanol on GABA-gated currents from Ltk cells expressing specific GABAA receptor subunits. Currents gated by 1.5 µM GABA (EC15) were measured in Ltk cells expressing 1, 2, and either 2S (Gamma2S) or 2L (Gamma2L) subunits for the GABAA receptor. Percentage of change in the GABA-gated current was plotted in the presence of varying concentrations of ethanol (25-300 mM). Control (cont) currents represent the percentage of change between two repeated presentations of GABA alone. There was no significant effect of ethanol on either subunit combination (P > 0.1). Each column represents data from 5 to 14 neurons.

Effect of Alphaxalone on GABA-Gated Currents. The lack of effect of ethanol on GABA-gated currents suggested that the acutely dissociated neurons might be in a state where drugs acting at allosteric modulatory sites were unable to enhance the effect of GABA. To test this hypothesis, we examined the effect of the neuroactive steroid anesthetic alphaxalone on GABA-gated currents. Figure 11 shows that alphaxalone robustly enhanced the GABA-gated currents to a similar degree in all brain regions examined. Additionally, alphaxalone enhanced the action of GABA in the Ltk cells. Thus, neurons in brain regions and cell types where the GABA-gated currents were unresponsive to ethanol were sensitive to alphaxalone.

View larger version (23K): [in this window] [in a new window]   Fig. 11.   Effect of alphaxalone on GABA-gated currents from several cell types. Alphaxalone enhanced GABA-gated currents from neurons acutely isolated from the substantia nigra (sn), lateral septum (ls), or cerebellar Purkinje neurons (pkj). Similar effects were obtained from Ltk cells expressing 1, 2 and 2 L (g2l) or 2S (g2S) GABAA receptor subunits. Significant linear trends were observed at each site or cell type, indicating a concentration-related effect of the alphaxalone. Each column represents data from 4 to 13 neurons.

Interaction between Ethanol and Alphaxalone. Because neuroactive steroids are not present in vitro it is possible that their absence might account for the lack of an effect of ethanol in vitro. In contrast to our previous report (Criswell et al., 1999), Fig. 12 shows that ethanol did not enhance the ability of alphaxalone to augment GABA-gated currents in the substantia nigra. Likewise, ethanol did not enhance the action of alphaxalone and actually decreased the effect of high concentrations of alphaxalone on GABA-gated currents from cerebellar Purkinje neurons.

View larger version (21K): [in this window] [in a new window]   Fig. 12.   Effects of ethanol on GABA-gated currents in the presence of alphaxalone. Concentration-response curves were compiled for the effect of ethanol (10-300 mM) on GABA-gated currents from substantia nigra (left) or cerebellar Purkinje neurons (right) in the presence of 30 to 300 nM alphaxalone (ax30-ax300). A, two-way analysis of variance failed to detect a significant effect of ethanol (P > 0.05) or a significant interaction between ethanol and alphaxalone concentration (P > 0.05) for the substantia nigra neurons. B, there was a significant interaction between alphaxalone concentration and ethanol concentration (P < 0.01) for cerebellar Purkinje neurons such that there was a concentration-related decrease (linear trend; P < 0.01) in GABA-gated currents with increasing concentration of ethanol at the 100 and 300 nM alphaxalone concentrations but no linear trend at the 30 nM alphaxalone concentration (P > 0.1). Each column represents data from 8 to 10 neurons.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Ethanol produced a concentration-dependent inhibition of NMDA-gated currents in dissociated neurons and in CHO cells expressing NMDA receptors. This is in agreement with previous work in cells cultured from cerebral cortex (Lovinger, 1995; Ming et al., 2001), striatum (Popp et al., 1998), HEK-293 cells (Lovinger, 1995), and neurons acutely dissociated from the medial septum (Grover et al., 1998). The concentration dependence of the effect of ethanol on NMDA-gated currents was similar between brain regions with the only statistically reliable differences representing the most sensitive (lateral septum) and least sensitive (thalamus) of the brain regions. This in vitro result is in direct contrast to earlier in vivo studies where ethanol decreased the effect of NMDA on neural activity from the substantia nigra but had no effect on responses from the lateral septum (Simson et al., 1993). Furthermore, systemic doses of ethanol as high as 1.5 g/kg did not alter the spontaneous firing rate of lateral septal neurons (Givens and Breese, 1990a). In contrast, ethanol inhibited the effect of NMDA at several other brain areas, including the substantia nigra, medial septum, and hippocampus (Simson et al., 1991, 1993; Yang et al., 1996).

One possible explanation for the in vivo data was that the neurons in which ethanol antagonized the action of NMDA contained a different set of subunits for the NMDA receptor than did neurons sensitive to the effects of ethanol. In this respect, Lovinger (1995) found a correlation between the appearance of the NR2B subunit and ifenprodil sensitivity during time in culture and ethanol inhibition of NMDA-gated currents. Additionally, Yang et al. (1996) found a correlation between the ability of ethanol and ifenprodil, an NMDA antagonist selective for NR2B subunit-containing receptors (Williams, 1993), to inhibit NMDA receptor function in vivo. However, a causal nature for this correlation between subunit composition or sensitivity to ifenprodil and sensitivity to ethanol has not been born out in experimental studies using transient expression systems (Blevins et al., 1997; Popp et al., 1998). Similarly, when the effects of ethanol on NMDA receptors containing either the NR2A or NR2B subunit transiently expressed in CHO cells were compared in the present study, there was no clear difference. These results support the position that the correlations between ifenprodil sensitivity and ethanol sensitivity described above are not causal in nature. However, because NMDA receptors expressed in CHO cells may experience an environment different from that found in neurons, we used the NR2B subunit-selective NMDA antagonist ifenprodil to determine the extent to which NMDA-gated currents were mediated by receptors containing the NR2B subunit in neurons from the lateral septum and substantia nigra.

In agreement with binding studies showing relatively low levels of binding of NR2B-selective ligands in the lateral septum (Mutel et al., 1998), ifenprodil blocked only 15% of the NMDA-gated current in the lateral septum but 43% of the NMDA-gated currents in the substantia nigra in the present investigation. To determine whether NR2B subunit-containing receptors mediated a disproportionate fraction of the effect of ethanol on NMDA-gated currents in neurons, we examined the ability of ethanol to inhibit NMDA-gated currents in the presence and absence of ifenprodil. When the effect of ethanol on NMDA-gated currents in the presence or absence of ifenprodil was examined in the substantia nigra, the ethanol produced a similar percentage of inhibition of the remaining current in the two conditions. Thus, for the substantia nigra, the ifenprodil-sensitive and ifenprodil-insensitive NMDA receptors had a similar sensitivity to ethanol. In contrast to the results from the substantia nigra, a different pattern of ethanol sensitivity was observed in the lateral septum where ethanol was significantly less effective in the presence of ifenprodil. This result would usually imply that the ifenprodil-insensitive receptors were also less sensitive to ethanol as we have observed in vivo (Yang et al., 1996). However, current carried by ifenprodil-sensitive receptors in the lateral septum represented only a small percentage (15%) of the total NMDA-gated current. Therefore, removal of this small pool of receptors by ifenprodil would not be expected to produce the magnitude of effect on ethanol sensitivity observed. One explanation for the apparent decrease in the effect of ethanol at neurons in the lateral septum in the presence of ifenprodil could be an interaction between ethanol and ifenprodil such that ifenprodil was less effective in blocking NMDA-gated currents in the presence of ethanol. The lack of such an interaction in CHO cells expressing combinations of NMDA subunits observed in the lateral septum argues against the above-mentioned hypothesis. A second possibility would be an interaction of ifenprodil with other neurotransmitter systems to indirectly act on the NMDA receptor (McCool and Lovinger, 1995). However, because the present study used acutely dissociated neurons in a defined medium (SES) without the presence of other neurotransmitters, this action of ifenprodil cannot explain its interaction with ethanol. The regional difference in the interaction between the effects of ethanol and ifenprodil remains to be explained. Nonetheless, the stark differences between the brain regional differences observed in vivo and the lack of differences in vitro suggest that variables other than the effectiveness of ethanol at the postsynaptic junction of the NMDA receptor play an important role in the effect of ethanol on neuronal activity.

As was the case for effects of ethanol on NMDA function, ethanol has been shown in vivo, to enhance the effect of GABA in some brain regions, but not others (Givens and Breese, 1990a,b; Criswell et al., 1995). Even within brain regions, some neurons show an enhancement of the effect of GABA by ethanol, whereas others do not (Criswell et al., 1995). At a behavioral level, microinjection of GABA agonists or antagonists into specific brain sites also enhances or depresses, respectively, the behavioral actions of ethanol but does not interact with ethanol when microinjected into other brain areas (McCown et al., 1986; Givens and Breese, 1990b). These observations are in contrast to the present study where ethanol was without effect on GABA-gated currents from neurons from several brain sites, over a wide range of concentrations and over a wide range of neuronal types. This lack of effect of ethanol in vitro is in agreement with a number of recent studies that have failed to observe effects of ethanol even at high concentrations (100-300 mM) (Sigel et al., 1993; Frye et al., 1994; Mori et al., 2000; Ming et al., 2001), but at odds with others (Reynolds et al., 1992; Weiner et al., 1994; Sapp and Yeh, 1998) and with the similarity of the behavioral effects of ethanol with those of agents acting at the GABA_A receptor (Frye et al., 1980; Liljequist and Engel, 1982; Martz et al., 1983). One possible reason for the lack of a reliable effect of ethanol on GABA-gated currents in vitro is that GABA_A receptors in acutely dissociated neurons are unable to respond to allosteric modulators. The robust response to alphaxalone by all cell types in the present study indicates that ability of these cells to respond to allosteric modulators remains intact. A second possibility is that the cells must be in a specific state to respond to ethanol and that is uncontrolled across studies. For example, Sapp and Yeh (1998) were able to see an effect of ethanol on GABA-gated currents from acutely dissociated cerebellar Purkinje neurons but found no effect of ethanol on HEK-293 cells expressing the same GABA_A subunits found in the cerebellar Purkinje neurons. Similarly, Mori et al. (2000) found a 25% enhancement of GABA-gated currents at 169 mM ethanol in cultured dorsal root ganglion cells, but HEK-293 cells expressing the 122L or 122S subunit combination required over 500 mM ethanol for a 25% enhancement. In cultured cortical neurons, they were unable to obtain a 25% enhancement of GABA-gated currents even at 1000 mM, the highest level tested. Using a slice preparation, others have found effects of ethanol on GABA-gated currents only after activation of protein kinase C (Weiner et al., 1994) or exposure to a GABA_B agonist (Allan et al., 1991) or antagonist (Wan et al., 1996).

We previously reported an effect of ethanol on GABA-gated currents from substantia nigra neurons in the presence of neuroactive steroids, but we (Criswell et al., 2000) and others (Hsiao et al., 2001) have been unable to reproduce that observation. In the present study, the neuroactive steroid alphaxalone was highly effective in enhancing GABA-gated currents from all cells tested but ethanol did not augment this effect in the substantia nigra. In the presence of alphaxalone, ethanol inhibited GABA-gated currents in cerebellar Purkinje neurons. This latter result is in agreement with in vivo studies where ethanol decreases the inhibition of Purkinje neuron firing by GABA in some neurons but not in others (Lin et al., 1993; Yang et al., 2000). As noted above, the neuroactive steroids reliably enhance GABA-gated currents in agreement with others (Majewska et al., 1986; Maitra and Reynolds, 1998; Criswell et al., 1999; Hsiao et al., 2001). Similarly, we have been able to demonstrate effects of volatile anesthetics where ethanol is ineffective under the same conditions (Ming et al., 2001), and there are numerous examples of enhancement of GABA-gated currents by benzodiazepine agonists in similar preparations (Frye et al., 1994; Criswell et al., 1997). Thus, the above-mentioned data indicate that some variable other than subunit composition, sensitivity to allosteric modulation, or the differential presence of a neuroactive steroid must underlie the brain regional differences in the effect of ethanol on GABA function observed in vivo.

In summary, the pattern of effects of ethanol on NMDA and GABA_A receptor function in vivo differed from that observed in vitro. The regional differences in the effect of ethanol observed in vivo are not explained by regional differences in the expression of specific combinations of NMDA or GABA_A receptor subunits. These differing results seen in vivo versus in vitro may be because the effects of ethanol observed in vivo are dependent upon presynaptic mechanisms not present in vitro. Alternatively, the postsynaptic receptor may be altered by a number of humeral factors present in vivo in a brain region-dependent manner that are absent in vitro. One such humeral factor, the presence or absence of neuroactive steroid modulation of the GABA_A receptor, did not alter the action of ethanol on GABA-gated currents. The mechanism allowing ifenprodil to alter the effect of ethanol on NMDA-gated currents from the lateral septum, but not the medial septum, is yet to be resolved. Definition of this unknown mechanism will likely provide insight into the brain regional differences in the effect of ethanol on NMDA and GABA function observed in vivo.