Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

There is considerable evidence that an action of ethanol on the -aminobutyric acid (GABA)_A receptor complex contributes to its pharmacological properties (Frye and Breese, 1982; Liljequist and Engel, 1982; Crews et al., 1996). Furthermore, biochemical studies demonstrated ethanol enhancement of GABA-stimulated chloride flux in a variety of preparations (Mehta and Ticku, 1988; Suzdak et al., 1986; Ticku et al., 1986; Reynolds et al., 1992). Electrophysiological investigations also supported the contention that ethanol potentiated GABA function (Nesteros, 1980; Aguayo, 1990; Givens and Breese, 1990a,b; Nishio and Narahashi, 1990; Simson et al., 1991; Proctor et al., 1992; Criswell et al., 1993; Freund et al., 1993; Sapp and Yeh, 1998).

In spite of the various positive electrophysiological reports of ethanol enhancement of GABA function, there are a number of reports from electrophysiological studies where ethanol did not have this action (Siggins et al., 1987; Givens and Breese, 1990b; White et al., 1990; Criswell et al., 1993; Sigel et al., 1993; Frye et al., 1994). One explanation for these disparate findings was that ethanol enhancement of responses to GABA was dependent upon brain region (Givens and Breese, 1990b; Breese et al., 1993; Criswell et al., 1993; Soldo et al., 1994). However, this view of variability of ethanol action between brain regions did not explain the disparate results obtained from recombinant GABA_A receptors (Wafford et al., 1991; Sapp and Yeh, 1998). Although ethanol was found to enhance GABA responses from a specific recombinant GABA_A-receptor subunit combination transfected into oocytes (Wafford et al. 1991; Wafford and Whiting, 1992), other investigators (Sigel et al., 1993; Sapp and Yeh, 1998) did not observe an effect of ethanol when this receptor subunit combination was studied in cell lines.

In unpublished studies from our laboratory, it was found that ethanol only rarely enhanced GABA responses from neurons dissociated from substantia nigra reticulata (SNR), a brain region where ethanol enhanced responses to GABA from the majority of neurons in vivo (Criswell et al., 1993, 1995). Based upon the contrast of our in vivo and in vitro results, we hypothesized that a factor present in vivo, but not available to isolated neurons in vitro, was important for a consistent effect of ethanol on GABA_A receptor function. Possible factors considered were those substances acting on one of the auxiliary sites influencing GABA_A receptor function (Sieghart, 1995). The neurosteroids are one group of endogenous compounds known to act on an auxiliary site on GABA_A receptors (Turner et al., 1989). Thus, the present study examined the action of the neurosteroid analog alphaxalone on the consistency and degree to which ethanol affected GABA-gated currents from dissociated SNR neurons.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Drugs. Alphaxalone (Research Biochemicals Inc., Natick, MA) and 3,5-tetrahydrodeoxycorticosterone (THDOC) (Research Biochemicals Inc.) were initially dissolved at a concentration of 10 mM in 100% dimethyl sulfoxide (DMSO) (Sigma Chemical Co., St. Louis, MO). This solution was diluted to 1 part per 1000 or less at the time of testing to attain the various concentrations of the neuroactive steroids. DMSO did not affect GABA-gated currents except for a slight inhibition at the highest concentration used (0.1% for 10 µM alphaxalone; N = 5, data not shown). Because alphaxalone alone gated a GABA-like current at concentrations of 3 µM, all interactions between alphaxalone and ethanol were tested at alphaxalone concentrations of 1 µM. GABA (5 mM; Sigma Chemical Co.) and ethanol (25, 50, or 100 mM; Aaper, Shelbyville, KY) were dissolved in a standard external solution (SES) [145 mM NaCl, 5 mM KCl, 10 mM HEPES, 2 mM CaCl₂, 1 mM MgCl₂, and 10 mM glucose (pH = 7.4; 340 mOsmol)].

Dissociation of Cells from SNR for Electrophysiological Recording. The basic procedure described by Oh et al. (1995) was used for dissociating neurons from the SNR. The SNR was dissected from 15- to 18-day-old rats. The ages chosen minimized the presence of neonatal forms of the GABA_A receptor. The excised tissue was immersed in a storage chamber containing SES. This mixture was held at 24°C, and bubbled with 100% O₂ for 60 min to allow stabilization of the neurons. Subsequently, the tissue was incubated for 30 min with protease XXIII from Aspergillus oryzae (3 mg/ml; Sigma Chemical Co.) at 24°C in the presence of 100% O₂ (Kaneda et al., 1988). The tissue was removed from the tube, washed, and subjected to gentle trituration to free neurons. The dissociated cells were then isolated from tissue debris, transferred to the recording chamber, and allowed to settle for ~10 min or until attached to the surface of the recording chamber. Response characteristics of the dissociated neurons were then defined with patch-clamp electrophysiology.

Electrophysiological Recording. After neurons settled onto the coverslip forming the bottom of the recording chamber, the chamber was bathed with SES flowing at 0.5 to 1.0 ml/min, which allowed debris to wash away. Neurons were viable for an ample time to record their electrophysiological responses to GABA in the presence and absence of ethanol and the neuroactive steroids. Recording from the SNR neurons was performed under voltage clamp in the whole cell-configuration with the use of an Axopatch 2d patch clamp amplifier (Criswell et al., 1997). Recording pipettes were fabricated from N 51A capillary glass (Drummond Scientific Co., Broomall, PA). The internal solution included 150 mM KCl, 3.1 mM MgCl₂, 15 mM HEPES, 2 mM K-ATP, 5 mM ethylene glycol bis(-aminoethyl ether)-N,N,N',N',-tetraacetic acid, 15 mM phosphocreatinine, and 50 U/ml creatinine phosphokinase at a pH of 7.4 and osmolality of 310. Inclusion of the last two items restores energy sources and increases recording time (Forscher and Oxford, 1985). Seals were formed on the neurons with electrodes having tip resistance of 2 to 4 M. Data were displayed on an oscilloscope, digitized at 20 samples/s, and stored on a computer. Recordings were performed at room temperature (22-24°C). GABA alone and varying concentrations of ethanol, neurosteroids, or both with GABA were applied for 4-s intervals by a U-tube placed 50 to 100 µm from the neuron that allowed rapid application and removal of the drugs (McGehee and Oxford, 1991). There was a minimum of 1 min between drug applications.

Statistical Evaluation. Logarithmic concentration-response curves for the enhancement of GABA-gated currents by alphaxalone were fit to a sigmoid curve with GraphPad Prism software (GraphPad, San Diego, CA). A two-way ANOVA was used to evaluate the interaction between the effect of ethanol and neurosteroids on GABA-gated currents. If a significant main effect was found, simple effects were tested with a Tukey honestly significant difference test. One data point that was more than 3 S.D.s from the mean of remaining data was excluded from the analysis. The effect of a single concentratlion of ethanol in the presence and absence of alphaxalone was evaluated with a paired t test. Regression analysis was used to determine the concentration-dependence of the effect of THDOC on ethanol enhancement of GABA-gated currents. Based upon previous reports (Frye et al., 1994; Criswell et al., 1997; Sapp and Yeh, 1998), an increase in the response to GABA in the presence of ethanol of 20% was considered a significant, positive change.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effect of Alphaxalone on Responses to GABA from SNR Neurons. Based upon previous work showing that the ED₅₀ value for the effect of GABA on SNR cells was ~10 µM (Criswell et al., 1997), a concentration of 5 µM GABA was used in the present study to allow enhancement by ethanol and neurosteroids. This GABA-gated current has been shown to be blocked by bicuculline, indicating that it is mediated by GABA_A receptors (Criswell et al., 1997). As shown in Fig. 1, the neuroactive steroid alphaxalone increased the response to 5 µM GABA in a concentration-dependent manner.

View larger version (13K): [in this window] [in a new window]   Fig. 1.   Concentration-effect curve for enhancement of the effect of GABA on SNR neurons by alphaxalone. Higher concentrations of alphaxalone were not tested due to an interaction with the DMSO solvent and gating of a GABA-like current by alphaxalone alone. Details concerning the use of alphaxalone are provided in Materials and Methods.

Effect of Ethanol on Responses to GABA from Dissociated SNR Neurons. When the action of 100 mM ethanol was tested on GABA-gated currents from SNR neurons, ethanol produced an enhancement of GABA of 20% in only 4 of 18 SNR neurons (Frye et al., 1994; Sapp and Yeh, 1998; this study). However, in this initial evaluation, the overall increase in GABA responsiveness by 100 mM ethanol from all neurons was not significant (8.1% ± 4.6%; P > .05). Figure 2 illustrates this degree of inconsistency observed for the effects of ethanol on responses to GABA from dissociated SNR neurons. Although the recording in Fig. 2 (left) illustrates the lack of effect of 100 mM ethanol on a GABA-induced current observed from the majority of SNR neurons (Fig. 3, left), the recording in Fig. 2 (right) depicts the rare enhancement of GABA responsiveness induced by ethanol.

View larger version (29K): [in this window] [in a new window]   Fig. 2.   Examples of the variable effects of ethanol on responses to GABA. The recording on the left shows the lack of an effect of ethanol (EtOH; 100 mM) on a response to GABA (CONT) from an SNR neuron. The recording on the right illustrates the rare action of ethanol to enhance a response to GABA. Pre, response to GABA before addition of EtOH; post, response to GABA after ethanol application.

View larger version (29K): [in this window] [in a new window]   Fig. 3.   Illustration of the effectiveness of alphaxalone to allow ethanol to enhance GABA function from an SNR neuron. The recording on the left shows the effect of ethanol (EtOH; 100 mM) on a response to GABA (CONT) from an SNR neuron where ethanol had no effect on the response to GABA (see Fig. 2). The recording on the right is from the same neuron in the presence of 100 nM alphaxalone (AX). The 25, 50, and 100 (right) refer to the millimolar concentration of ethanol. See Fig. 4 for other group findings. Note the lack of effect of 100 mM ethanol on GABA function (left) compared with this concentration of ethanol on GABA responsiveness when alphaxalone was present (right).

Effect of Alphaxalone on Ethanol Enhancement of GABA-Induced Currents from Dissociated SNR Neurons. In contrast to the variable weak effect ethanol exerted on responses to GABA in the absence of alphaxalone (Fig. 2), ethanol (50 and 100 mM) had a potent effect on responses to GABA in the presence of alphaxalone (30 and 100 nM) from 12 of the 14 dissociated SNR neurons in this sample (Figs. 3 and 4). Although the mean percentage of enhancement of GABA by 100 mM ethanol alone from this particular group of neurons was 6.5% ± 5.1%, an increase to 63.4% ± 11.7% was observed in the presence of 30 or 100 nM alphaxalone (t₁₃ = 4.55; P < .01). An example of the extent to which 100 nM alphaxalone affected the ability of varying concentrations of ethanol to enhance GABA responsiveness from a neuron relatively insensitive to ethanol in the absence of alphaxalone is presented in Fig. 3.

View larger version (43K): [in this window] [in a new window]   Fig. 4.   Effect of varying concentrations of alphaxalone on the action of 25, 50, and 100 mM ethanol (EtOH) on GABA-gated currents. The various concentrations of alphaxalone (10, 30, 100, 300, and 1000 nM) tested against each of the ethanol concentrations are shown on the x-axis. Each bar represents mean ± S.E. of data from four to seven neurons. The percentage of enhancement by ethanol is normalized to the effect of 5 µM GABA in the presence of the concentration of alphaxalone for that group. There was a significant main effect of alphaxalone concentration (F5,92 = 8.12; P < .01) and of ethanol concentration (F2,92 = 7.76; P < .01). Tukey honestly significant difference tests showed that the effect of ethanol at 30, 100, and 300 nM alphaxalone differed from the effect of ethanol alone (P < .05) and asterisks mark significant differences between individual concentrations of ethanol in the presence of alphaxalone and the corresponding ethanol concentration in the absence of alphaxalone (P < .05; least-significant difference test).

Effect of THDOC on Ability of Ethanol to Enhance Responses to GABA from Dissociated SNR Neurons. Because the neurosteroid analog alphaxalone is not endogenous, we examined the effect of the endogenous compound THDOC (Majewska et al., 1986) to determine whether this neurosteroid could mimic the effects of alphaxalone. Figure 5 shows that ethanol consistently enhanced GABA-gated currents in the presence of 100 nM THDOC, just as seen in the presence of alphaxalone.

View larger version (17K): [in this window] [in a new window]   Fig. 5.   Effect of THDOC on the action of ethanol to enhance GABA-gated current. In the presence of THDOC (100 nM), ethanol (25, 50, and 100 mM) enhanced the effect of GABA in a concentration-related fashion as indicated by a significant linear trend (F1,19 = 21.33; P < .01). Asterisks indicate a significant effect of ethanol in the presence of THDOC compared with the effect of 100 mM ethanol in the absence of THDOC (CONT) (P < .05; Tukey honestly significant difference test). Values for the effect of ethanol in the presence of THDOC are normalized to the effect of 5 µM GABA in the presence of THDOC, whereas the effect of ethanol in the absence of THDOC is normalized to the effect of 5 µM GABA alone.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Previous in vivo electrophysiological investigations showed that ethanol consistently enhanced responses to GABA from SNR neurons (Criswell et al., 1993, 1995). Therefore, it would have been expected that dissociated SNR neurons would exhibit a consistent ethanol augmentation of GABA function. The present in vitro studies confirmed a previous impression that ethanol only rarely enhanced GABA function from dissociated SNR cells. Because the neurons for the in vitro and in vivo investigations were from the same brain region, brain region differences could not be a basis of these disparate results (Givens and Breese, 1990b; Criswell et al., 1993; Soldo et al., 1994). Early work indicated that brain regions where ethanol enhanced the effect of GABA (Givens and Breese, 1990b; Breese et al., 1993) also exhibited high levels of zolpidem binding (Duncan et al., 1995). Because zolpidem binding defines the presence of a type 1 benzodiazepine (BZD) receptor, it was hypothesized that this specific GABA_A receptor subtype was ethanol-sensitive (Breese et al., 1993; Criswell et al., 1993). This view was strengthened when Criswell et al. (1995) showed that virtually all SNR neurons responsive to ethanol in vivo were sensitive to zolpidem enhancement of responses to GABA. Subsequent characterization of mRNAs in acutely dissociated SNR neurons sensitive to zolpidem provided further evidence that the majority of this cell type contained a type 1 BZD receptor (Criswell et al., 1997). However, the inconsistent action of ethanol on GABA function from dissociated SNR neurons in the present study provides convincing evidence that specific receptor structure alone is not sufficient to explain ethanol's selective in vivo interaction with GABA_A receptors in the SNR (Criswell et al., 1993, 1995).

A potential difference between the in vivo studies in the SNR and those with acutely dissociated SNR neurons was the milieu in which the studies were performed. This view in turn suggested that a required "factor" for ethanol enhancement of GABA responsiveness that was present in vivo was not present in vitro. Based upon the presence of auxiliary sites on the GABA_A receptor complex that influence GABA_A receptor function (Sieghart, 1995), it was hypothesized that an endogenous compound acting on one of these auxiliary sites in vivo would have to be present in vitro for ethanol to have maximal effectiveness on responses to GABA from dissociated neurons. Neurosteroids were likely candidates because they act on an auxiliary site (Turner et al., 1989) and are known to exist in brain (Baulieu, 1991). Alphaxalone, a synthetic nonmetabolized neurosteroid analog (Maitra and Reynolds, 1998), was chosen to test this view.

Alphaxalone itself had a significant, concentration-related effect on GABA_A receptor function, in accord with previous findings (Maitra and Reynolds, 1998). Thus, the neurosteroid site was both present and functional on the acutely dissociated SNR neurons. Moreover, in the presence of 30 and 100 nM alphaxalone, the action of ethanol on GABA responsiveness from the SNR neurons was robust and consistent, providing a distinct contrast with the small, inconsistent effects of ethanol on GABA-gated currents from this cell type in the absence of alphaxalone. To determine whether ethanol enhancement of GABA-gated currents in the presence of alphaxalone was unique to this neuroactive steroid, the effect of the neurosteroid THDOC (Majewska et al., 1986; Rupprecht et al., 1996) also was tested. Just as seen with alphaxalone, the simultaneous presence of THDOC with ethanol resulted in a consistent, concentration-related effect of ethanol on GABA responsiveness of cells that were relatively insensitive to ethanol before applying the THDOC. Consequently, these observations are consistent with the hypothesis that activation of the neurosteroid site on the GABA_A receptor is important for ethanol enhancement of GABA from dissociated SNR neurons. Furthermore, the absence of any neural input to these dissociated neurons clearly indicates that the ethanol-neurosteroid interaction is on postsynaptic GABA_A receptor function.

The exact mechanism by which this interaction between ethanol and neurosteroid occurs has yet to be determined. For example, it is unknown whether the neurosteroid is enhancing the action of ethanol on responses to GABA or whether the ethanol is enhancing the action of the neurosteroid to affect GABA responsiveness. Nonetheless, results in the present study may be relevant to this issue. As shown in Fig. 4, the effect of 1 µM alphaxalone on GABA responsiveness was not enhanced by ethanol, even though this concentration of alphaxalone alone produced <50% maximal response (i.e., a ceiling effect on GABA responsiveness has not been reached at this concentration of alphaxalone). Consequently, it seems unlikely that ethanol is enhancing the action of the neurosteroid on the response to GABA; otherwise, the response to GABA in the presence of ethanol could have been expected to be enhanced at 1 µM alphaxalone.

Although the results from the present investigation provide a possible basis for the lack of effect of ethanol on the majority of dissociated SNR neurons, several questions are raised by the present findings. For example, in spite of the impressive effects of the neurosteroid-ethanol interaction on GABA responsiveness observed from isolated SNR neurons at 50 and 100 mM ethanol, it is bothersome that GABA responses to 25 mM ethanol were not significantly affected by the neuroactive steroids. A dose of ethanol in vivo producing a 25 mM blood concentration (0.11%) would produce a person who is legally intoxicated and thus would be expected to have a potent effect on GABA function. A possible explanation could be that the age of the animal from which these neurons were isolated contributes to this paradox because it is known that young animals are relatively insensitive to ethanol's actions (Fang et al., 1997). Another unknown is the basis of ethanol enhancing GABA responses from a small number of SNR neurons in the absence of a neuroactive steroid. In another investigation, Sapp and Yeh (1998) found a consistent ethanol augmentation of GABA by 39% from 15 of 18 dissociated cerebellar Purkinje neurons from immature rats that are thought to contain a zolpidem-sensitive (type 1 BZD) receptor. An explanation could be that the subunit composition or the stoichiometry of the GABA_A receptor forming a type 1 BZD receptor subtype on the specific neurons sensitive to ethanol differs from those present in the majority of SNR neurons insensitive to ethanol (Sieghart, 1995). Additional work is required to resolve these uncertainties.

In addition to the differing effects of ethanol on GABA responses from neurons, there is controversy concerning the effect of ethanol on GABA responsiveness from recombinant type 1 BZD receptors. For example, Wafford et al. (1991) and Wafford and Whiting (1992) reported an ethanol enhancement of GABA responsiveness from an expressed recombinant receptor in oocytes with characteristics of a type 1 BZD receptor. However, there are now reports that ethanol is without an effect on GABA-gated currents from cell lines with recombinant type 1 BZD receptors (Sigel et al., 1993; Sapp and Yeh, 1998). An obvious question to be raised from these latter findings is whether the recombinant GABA_A receptors insensitive to ethanol enhancement of GABA would become sensitive in the presence of a neurosteroid. Additional experiments will be necessary to resolve this question.

In summary, the present results demonstrated that the neurosteroid analog alphaxalone and the neurosteroid THDOC produced a robust and consistent ethanol enhancement of GABA function from SNR neurons that were otherwise inconsistently affected by ethanol in the absence of these compounds. Thus, it is possible to speculate that a neurosteroid-ethanol interaction may prove crucial to the maintenance of ethanol's action on GABA_A receptors in vivo as well as in vitro. In this regard, an implication of the present work would be that the regional selectivity for ethanol affecting GABA responsiveness in vivo (Givens and Breese, 1990b; Criswell et al., 1993) may be predicated on a robust, regionally specific neurosteroid presence or a neurosteroid potency difference between GABA_A receptor subtypes in various brain regions (Nguyen et al., 1995). With these possibilities, the stage is set for future investigations to search for an integrated interpretation of neurosteroid-ethanol interactions on GABA_A receptor function. Finally, the demonstration of a reliable ethanol action on GABA function in the presence of neurosteroid is of practical importance. For example, the proposed role of differing GABA_A receptor subtypes in the selectivity of ethanol to affect GABA responses in vivo in various brain regions can now be explored and resolved in vitro (Breese et al., 1993; Criswell et al., 1993).