Introduction

Top Abstract Introduction Methods Results Discussion References

Chronic ethanol consumption alters the function of GABA_A receptors in laboratory animals and in humans (see Morrow, 1995). In addition to activation by GABA and related agonists (e.g., muscimol), GABA_A receptor function can be modulated by many other agents, including ethanol, benzodiazepines, barbiturates and neurosteroids. Chronic ethanol treatment abolishes the ability of ethanol to acutely potentiate the function of GABA_A receptors in some preparations and provides a plausible mechanism for tolerance to at least some of the actions of ethanol (Buck and Harris, 1991). In addition, chronic ethanol exposure reduces the action of benzodiazepine agonists in animals as well as humans (Buck and Harris, 1991; Volkow et al., 1993). There also is evidence that chronic ethanol consumption enhances the action of benzodiazepine inverse agonists and neurosteroids and, in some studies, inhibits the actions of barbiturates (Buck and Harris, 1990a, 1990b; Devaud et al., 1996; Mehta and Ticku, 1989; Morrow et al., 1988). These diverse changes in GABA_A receptor function may be clinically important for ethanol tolerance and dependence as well as for cross-tolerance to other depressant drugs, such as benzodiazepines.

An important question is what are the cellular mechanisms responsible for these changes in receptor function? In many areas of biology, understanding of receptor regulation has required a cell culture system in which heterogeneity is reduced by the use of a clonal population of cells with defined receptors. Cells stably transfected with GABA_A receptor subunits have proved to be useful for understanding the actions of drugs on receptors of defined composition (Hadingham et al., 1992, 1993, 1995, 1996; Horne et al., 1992, 1993; Wong et al., 1994). We recently applied this approach to study the regulation of GABA_A receptors by ethanol and other drugs using cells stably transfected with receptor subunits (Klein et al., 1994, 1995a, 1995b). We used fibroblast-like cells expressing bovine alpha-1 beta-1 gamma-2L GABA_A receptor subunits and found that chronic (4 day) ethanol treatment "uncoupled" the GABA and benzodiazepine receptors; that is, it reduced the ability of GABA to enhance benzodiazepine binding and reduced the ability of benzodiazepines to enhance GABA receptor function (Klein et al., 1995a). This study showed the feasibility of using stably transfected cells to study regulation of GABA_A receptors by chronic ethanol treatments and raised a number of possibilities for further research. A particular advantage of stably transfected cells is that they allow study of receptors with defined subunit composition. In addition, they provide the opportunity to determine whether changes in gene expression or changes in post-translational processing are responsible for receptor adaptation produced by chronic ethanol treatments (Klein et al., 1995a, 1995b).

The present series of studies was carried out to answer the following questions: (1) Are the chronic effects of ethanol dependent on the subunit composition of the receptor? The previous study tested only a single subunit combination, and it was of interest to test other subunits combinations that are likely to exist in brain. In particular, we found that only cells expressing receptors containing a gamma-2L subunit were potentiated by acute exposure to low concentrations of ethanol (Harris et al., 1995b, 1997), and it was of interest to determine whether acute sensitivity to ethanol was required for receptor regulation by chronic exposure. (2) Does chronic ethanol exposure affect the ability of ethanol, benzodiazepine inverse agonists, neurosteroids or barbiturates to modulate receptor function? The previous study tested only a GABA_A agonist (muscimol) and a benzodiazepine agonist (flunitrazepam), and it was of interest to test other drugs that have been examined in animal studies. (3) What are the characteristics of changes produced by ethanol exposure (e.g., concentration and time dependence)? In the present study, we evaluate the ethanol concentration dependence for chronic exposure as well as the time course for acquisition (and, in some cases, reversal) of receptor changes.

For this study, we used cells stably transfected with six different combinations of GABA_A receptor subunits and measured receptor function by a ³⁶Cl flux assay. These subunit combinations were chosen to represent receptors that are likely to exist in the brain (McKernan and Whiting, 1996; Whiting et al., 1995). We recently characterized the acute effects of muscimol, ethanol, flunitrazepam and pentobarbital on these cells by the ³⁶Cl flux technique (Harris et al., 1995b, 1997). The chronic alcohol treatment was based on our previous study (Klein et al., 1995b) and initially used continuous exposure to 100 mM alcohol for 1 to 4 days to test receptor function. We then varied the duration of ethanol exposure as well as the concentration of ethanol to determine the time course of receptor regulation as well as the sensitivity to chronic ethanol.

	Methods

Top Abstract Introduction Methods Results Discussion References

Materials. Muscimol and DMCM were purchased from Research Biochemicals (Natick, MA), flunitrazepam and pentobarbital were from Sigma Chemical (St. Louis, MO) and pregnanolone was from Steraloids (Wilton, NH). GF109203X (bisindolymaleimide I, HCl) was from Calbiochem (La Jolla, CA), and ethanol was purchased from Aaper Alcohol and Chemical (Shelbyville, KY). Sulfosuccimido-NHS-biotin and monomeric avidin beads were from Pierce (Rockford, IL). Anti--GABA_A receptor monoclonal antibody (clone bd 24) and enhanced chemiluminescence kit were from Boehringer-Mannheim Biochemicals (Indianapolis, IN). All other chemicals used were of reagent grade.

Cell culture conditions. Stable transfection of mouse L (tk) cells with human alpha-1 beta-2 gamma-2L, alpha-1 beta-2 gamma-2S, alpha-5 beta-3 gamma-3, alpha-1 beta-3 gamma-2S and bovine alpha-1 beta-1 subunits were carried out as previously described for human alpha-6 beta-3 gamma-2S (Hadingham et al., 1996) GABA_A receptor subunits. Expression was controlled by a dexamethasone-sensitive promoter as described by Hadingham et al. (1992).

Chronic ethanol treatment. Dexamethasone-induced cells growing on coverslips were exposed to ethanol for up to 4 days. Ethanol (100 mM, unless stated otherwise) was added 3 to 4 days after the addition of dexamethasone. Media, including dexamethasone, and ethanol were changed after 2 days of ethanol exposure. To prevent evaporation of ethanol in the incubator, six-well plates were placed in Ziploc bags with a beaker of 400 ml of ethanol at the same concentration as used for the cells (Chandler et al., 1993). Under these conditions, the ethanol concentration is unchanged after 2 days (Klein et al., 1995b). Control cultures were handled in the same manner.

Chloride flux measurements. For measurement of ³⁶Cl uptake, cells were washed twice (3 min/wash) in wash buffer consisting of 136 mM NaCl, 5.4 mM KCl, 1.4 mM MgCl₂, 1.2 mM CaCl₂ 1 mM NaH₂PO₄ and 20 mM 4-(2-hydroxyethyl)-1-piperazine ethanesulfonic acid, adjusted to pH 7.4 with Tris base. This wash solution was at room temperature and contained the same concentration of ethanol as was used for the chronic treatment (e.g., none for control cells and 10-200 mM, usually 100 mM, for chronic ethanol).

Quantification of external GABA_A receptors. Cells stably expressing alpha-1 beta-2 gamma-2L GABA_A subunits were grown and treated with dexamethasone and treated with or without 100 mM ethanol for 2, 22 or 45 hr. Surface membrane proteins were labeled with the membrane impermeant biotin derivative, sulfosuccimido-NHS-biotin, and precipitated with monomeric avidin beads as described elsewhere (Qian et al., 1997). Then, 40 µl of total cell lysate and eluates after avidin bead precipitation was separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (10%), blotted with 1 µg/ml anti--GABA_A receptor monoclonal antibody (clone bd 24) and probed with 1 µg/ml goat anti-rabbit horseradish peroxidase-conjugated secondary antibody. Immunoreactive bands were visualized by enhanced chemiluminescence on X-ray film, scanned and quantified with Image PC Beta-1 computer program (Scion Corp, Frederick, MD).

Statistical analyses. Concentration-response curves were fitted to a sigmoid function using GraphPAD Inplot software (GraphPAD Software, San Diego, CA). The effects of single concentrations of drugs were evaluated by a two-tailed Student's t test to assess for statistical significance. The effects of ethanol exposure on actions of multiple concentrations of drugs were determined by analysis of variance for repeated measures. The effects of chronic ethanol treatment on muscimol action and flunitrazepam enhancement of muscimol action are presented as percent reduction, which was calculated as [1  (response of ethanol-treated cells)/(response of control cells)] × 100.

	Results

Top Abstract Introduction Methods Results Discussion References

Ethanol treatment reduced muscimol action. Cells were treated with ethanol (100 mM) for 24 hr, the ethanol was removed and the muscimol-stimulated uptake of ³⁶Cl was measured as an indicator of GABA_A receptor function. This ethanol treatment reduced the maximal effect (E_max) of muscimol on alpha-1 beta-2 gamma-2S receptors and also produced a small reduction in the muscimol EC₅₀ values (figs. 1 and 2). Chronic ethanol treatment reduced the muscimol E_max values for cells expressing human alpha-1 beta-2 gamma-2L, alpha-1 beta-2 gamma-2S, alpha-5 beta-3 gamma-3, alpha-1 beta-3 gamma-2S, alpha-6 beta-3 gamma-2S and bovine alpha-1 beta-1 receptors and reduced the EC₅₀ values in cells expressing alpha-1 beta-2 gamma-2S, alpha-1 beta-3 gamma-2S and alpha-5 beta-3 gamma-3 subunits (fig. 2). It should be noted that chronic ethanol treatment did not change the muscimol EC₅₀ values of cells expressing the alpha-1 beta-2 gamma-2L subunits, and these cells were used for most of the subsequent experiments.

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Chronic ethanol treatment inhibits muscimol-stimulated 36Cl flux. Cells expressing alpha-1 beta-2 gamma-2S receptor subunits were treated with or without (Control) 100 mM ethanol for 24 hr, and muscimol-stimulated 36Cl uptake measured in the absence of ethanol. Statistical analysis of effects of chronic ethanol on the muscimol EC50 and Emax are given in figure 2. Values are mean ± S.E.M. (n = 6).

View larger version (31K): [in this window] [in a new window]   Fig. 2.   Effects of chronic ethanol treatment on muscimol-stimulated 36Cl flux in cells expressing six different subunit combinations. Cells were treated with or without (Control) 100 mM ethanol for 24 hr, and muscimol concentration-response curves were determined, as shown in figure 1. Top, maximal uptake. Bottom, concentration of muscimol producing a half-maximal stimulation of 36Cl uptake. Values are mean ± S.E.M. (n = 6-24). *Significant effect of chronic ethanol, P < .05; ** P < .01; *** P < .001, t test.

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Inhibition of muscimol action by chronic ethanol showing time course for acquisition and reversal and ethanol sensitivity. Top, cells expressing alpha-1 beta-2 gamma-2L receptors were treated for various lengths of time (x axis) with 100 mM ethanol, and muscimol-stimulated 36Cl flux was measured. Middle, after exposure to 100 mM ethanol for 24 hr, cells were placed in medium without ethanol for the indicated times, and muscimol-stimulated 36Cl uptake was measured. Bottom, cells were exposed to varying concentrations of ethanol for 24 hr before assay of 36Cl flux. Solid line was fit by linear regression. In all experiments, the concentration of muscimol was 10 µM. Values represent the percent reduction in muscimol action produced by chronic ethanol and are given as mean ± S.E.M. (n = 12-24). Values derived from these curves are summarized in table 1.

Ethanol treatment abolished ethanol action. Ethanol enhances muscimol-stimulated ³⁶Cl uptake in cells expressing a gamma-2L subunit together with alpha and beta subunits (Harris et al., 1995b, 1997). To determine whether chronic ethanol exposure produced tolerance to this action of ethanol, we treated cells expressing alpha-1 beta-2 gamma-2L subunits with 100 mM ethanol for different lengths of time and tested the ability of 50 mM ethanol to enhance muscimol action. Exposure to ethanol for 15 min to 48 hr abolished the ability of subsequent exposure to ethanol to enhance muscimol action (fig. 4, top). This tolerance occurred rapidly and was half-maximal in 10 min. Using a 4-hr exposure time, the concentration of ethanol was varied, and a concentration as low as 25 mM was found to produce almost complete tolerance (fig. 4, bottom).

View larger version (18K): [in this window] [in a new window]   Fig. 4.   Chronic ethanol treatment produces tolerance to acute actions of ethanol. Top, cells expressing alpha-1 beta-2 gamma-2L receptors were treated with 100 mM ethanol for varying lengths of time (x axis), and the ability of acute exposure to ethanol (50 mM) to enhance muscimol-stimulated 36Cl flux was measured. Bottom, cells were treated with varying concentrations of ethanol (x axis) for 4 hr, and the ability of ethanol to acutely enhance muscimol action was determined. Values are mean ± S.E.M. (n = 14-24). Values calculated from these results are summarized in table 1. The concentration of muscimol was 0.5 µM.

Ethanol treatment reduced flunitrazepam, DMCM and pregnanolone action. The ability of flunitrazepam to potentiate the action of a low concentration of muscimol was studied in cells (alpha-1 beta-2 gamma-2L ) treated with 100 mM ethanol for 4 days. This ethanol treatment reduced the maximal potentiation of flunitrazepam (fig. 5, top; see figure legend for statistics). A shorter exposure to ethanol (24 hr) was used to compare effects on cells expressing alpha-1 beta-2 gamma-2L or alpha-1 beta-2 gamma-2S receptors. A similar reduction in the effect of flunitrazepam (using a maximally effective concentration of flunitrazepam) was observed with both cell lines (fig. 5, bottom).

View larger version (22K): [in this window] [in a new window]   Fig. 5.   Chronic ethanol treatment reduces the action of flunitrazepam. Top, cells expressing alpha-1 beta-2 gamma-2L subunits were treated with ethanol (100 mM) for 4 days, and the ability of flunitrazepam to enhance muscimol action was measured. Values calculated from these curves were Emax (nmol/mg of protein) = 31 ± 3 (control), 22 ± 1 (ethanol), P < .01; EC50 (nM) = 639 ± 117 (control), 303 ± 55 (ethanol), P < .02 (n = 6-8). Bottom, comparison of chronic ethanol treatment (100 mM, 24 hr) on alpha-1 beta-2 gamma-2L and alpha-1 beta-2 gamma-2S receptors. Significant inhibition of flunitrazepam action, * P < .05; *** P < .001. Values are mean ± S.E.M. (n = 8-10). The concentration of muscimol was 0.5 µM, and the concentration of flunitrazepam was 10 µM. Changes of flunitrazepam action produced by ethanol treatments are summarized in table 1.

View larger version (23K): [in this window] [in a new window]   Fig. 6.   Chronic ethanol treatment reduces the action of DMCM and pregnanolone. Cells expressing alpha-1 beta-2 gamma-2L subunits were treated with or without (Control) ethanol (100 mM) for 24 hr and the ability of DMCM (top) to inhibit or pregnanolone (bottom) to enhance muscimol action was measured. Values calculated from these curves are as follows: DMCM, Emax = 37 ± 3 (control), 25 ± 1 (ethanol), P < .01; EC50 = 10 ± 3 (control), 9 ± 2 (ethanol); pregnanolone, Emax = 36 ± 2 (control), 27 ± 1 (ethanol), P < .01; EC50 = 0.7 ± 0.1 (control), 0.5 ± 0.1 (ethanol). Values are mean ± S.E.M. (n = 6-12). The concentration of muscimol was 2 µM for DMCM and 0.5 µM for pregnanolone.

Ethanol treatment and actions of pentobarbital and zinc. Pentobarbital enhancement of muscimol action was studied in cells expressing alpha-1 beta-2 gamma-2L receptors. Pentobarbital produced a biphasic action with potentiation of responses with concentrations of 10 to 100 µM and inhibition at 300 µM (fig. 7, top). Treatment with ethanol for 24 hr (100 mM) did not alter the actions of pentobarbital. Pentobarbital also produces a direct activation of GABA_A receptors, and this effect is most pronounced for receptors containing the alpha-6 subunit (Thompson et al., 1996). We studied this direct action of pentobarbital using cells expressing alpha-6 beta-3 gamma-2S receptors. Concentrations of 100 to 3000 µM pentobarbital directly activated ³⁶Cl uptake by these cells, and this action was enhanced by chronic ethanol treatment (fig. 7, bottom). However, this enhancement is quite small and unlikely to be functionally important.

View larger version (22K): [in this window] [in a new window]   Fig. 7.   Chronic ethanol treatment and the direct and indirect actions of pentobarbital. Top, cells expressing alpha-1 beta-2 gamma-2L subunits were treated with ethanol (100 mM, 24 hr), and the ability of pentobarbital to potentiate muscimol action was determined. The concentration of muscimol was 0.5 µM. There was no significant effect of ethanol (F = 1.9, d.f. = 5.1, analysis of variance for repeated measures). Bottom, cells expressing alpha-6 beta-3 gamma-2S subunits were treated with ethanol (100 M, 24 hr), and the ability of pentobarbital to directly activate 36Cl flux was determined. Ethanol treatment slightly enhanced pentobarbital action (F = 13, d.f. = 5.1; P < .005, analysis of variance for repeated measures; significant difference in maximal effect, t = 2.4, P < .05). Values are mean ± S.E.M. (n = 6).

View larger version (19K): [in this window] [in a new window]   Fig. 8.   Chronic ethanol treatment does not alter the action of zinc. Cells expressing alpha-1 beta-2 gamma-2L subunits were treated with ethanol (100 mM for 24 hr), and the ability of zinc to inhibit muscimol action was determined. Values are mean ± S.E.M. (n = 6). Concentration of muscimol was 2 µM.

Effect of ethanol on surface GABA_A receptor levels. To determine whether the effects of ethanol on GABA_A receptor pharmacology were due to a change in levels of surface receptors, we performed cell surface biotinylation experiments with the membrane-impermeable reagent sulfosuccimido-NHS-biotin. We measured total and external GABA_A receptor alpha-1 subunit immunoreactivity in Western blots of cells stably expressing human alpha-1 beta-2 gamma-2L subunits. To determine levels of external GABA_A receptor alpha-1, immunoreactivity was quantified after precipitation of surface biotinylated proteins with monomeric avidin beads. These experiments demonstrated that 50% of the GABA_A receptors are located in the cell surface (fig. 9). Moreover, treatment with 200 mM ethanol for 2 hr or with 100 mM ethanol for 22 or 45 hr did not significantly change the levels of surface receptors (fig. 9; P > .2 by analysis of variance). In addition, immunofluorescence analysis of surface receptor density of intact cells on coverslips did not detect any effects of chronic ethanol exposure (data not shown).

View larger version (19K): [in this window] [in a new window]   Fig. 9.   Ethanol does not affect the levels of surface GABAA receptors. Immunoblots of total and surface biotinylated-avidin precipitated lysate of cells expressing alpha-1 beta-2 gamma-2L subunits were probed with anti-alpha-1 GABAA receptor subunit antibody and quantified densitometrically. Surface receptor values are expressed as a percent of the total receptor immunoreactivity. Each bar represents the average ± S.E.M. of control cells (n = 8), cells treated with 200 mM ethanol for 2 hr (n = 4) and cells treated with 100 mM ethanol for 22 hr (n = 3) or 45 hr (n = 3). Ethanol did not significantly change the levels of surface GABAA receptors (P > .2 by analysis of variance).

Effects of a PKC inhibitor. Because the receptor adaptation produced by chronic ethanol exposure is most likely post-translational (see Discussion) and PKC is known to modulate the function and drug sensitivity of the GABA_A receptor (Moss and Smart, 1996), we asked whether an inhibitor of PKC would prevent the reduction in muscimol action that results from chronic ethanol exposure. In these studies, cells expressing alpha-1 beta-2 gamma-2S receptors were incubated with 200 nM of the selective PKC inhibitor GF109203X (Toullec et al., 1991) for 30 min before the addition of ethanol (100 mM). Controls were treated with GF109203X alone. Treatment with GF109203X with or without ethanol was continued for 48 hr, and muscimol-stimulated ³⁶Cl uptake was measured. Treatment with GF109203X alone did not alter the EC₅₀ or E_max value of muscimol action, and treatment with GF109203X plus ethanol resulted in a 30% reduction in the E_max with no change in the EC₅₀, just as was seen with cells treated with ethanol alone. Thus, GF109203X was not able to block the effect of chronic ethanol treatment on muscimol action.

	Discussion

Top Abstract Introduction Methods Results Discussion References

Chronic ethanol exposure produced changes in almost every aspect of GABA_A receptor function that was examined in this study. The three functions that were studied in detail (muscimol action and ethanol and flunitrazepam modulation) showed different sensitivities to chronic ethanol exposure and different time courses of acquisition and reversal of receptor function. In particular, development of tolerance to ethanol occurred quite rapidly (within minutes) and was the most sensitive in terms of concentrations of ethanol required for the chronic treatment (Table 1). The differences in ethanol sensitivity and time courses suggest multiple mechanisms are required to account for effects of chronic ethanol treatments on different aspects of GABA_A receptor function. It is important to note that a previous study showed that ethanol treatments did not change receptor density or the levels of mRNA or protein for GABA_A receptor subunits in a different line of stably transfected cells (Klein et al., 1995b). However, that study did not address the possibility that ethanol exposure affected receptor trafficking and altered the surface receptor density without affecting total receptor expression. In the present study, we separated surface and internal GABA_A receptors and found that chronic ethanol treatments did not affect surface receptor density. Thus, it is unlikely that ethanol treatments alter the expression of the GABA_A receptor subunits. Some of our results (e.g., decreased action of flunitrazepam and DMCM) could be accounted for by a loss of the gamma-2 subunit; however, zinc inhibition of GABA_A receptor function is very sensitive to the presence of this subunit (Smart et al., 1991), and we found that chronic ethanol treatment did not alter the action of zinc. Furthermore, these cells contain only one subunit of each type (e.g., alpha, beta and gamma), and there is no possibility for altering receptor function through substitution of different alpha, beta or gamma subunits. Chronic ethanol treatment alters the expression of specific receptor subunits in brain, and subunit substitution has been proposed as the mechanism for regulation of receptor function in vivo (Devaud et al., 1995; Mhatre and Ticku, 1992; Morrow, 1995). However, it is interesting to note that many of the changes in GABA_A receptor function observed after ethanol treatment in vivo are also noted in the present study of stably transfected cells. For example, enhancement of muscimol action by ethanol is rapidly (within 5 min) and completely eliminated by ethanol treatment in vivo (Allan and Harris, 1987; Morrow et al., 1988), whereas the reduction of flunitrazepam action produced by chronic ethanol consumption in vivo required longer treatment and was only a partial reduction (30%) (Buck and Harris, 1990b). In addition, the reduction of muscimol action was found only in rats with high blood ethanol levels, and this reduction was only partial (Morrow et al., 1988). Also, chronic ethanol treatments in vivo do not change the density of flunitrazepam binding sites in brain (Buck and Harris, 1990a, 1990b). However, there are several changes in receptor function that are different in vivo and in vitro. For example, we found that actions of pregnanolone and DMCM were reduced by chronic ethanol treatment of transfected cells, but ethanol consumption in vivo increased the action of these two drugs (Buck and Harris, 1990a, 1990b; Devaud et al., 1996). Perhaps the observed changes in subunit expression that occur in vivo are important for some of the changes in receptor function seen in vivo, such as enhancement of pregnanolone and DMCM action, but may not be for the changes in receptor function that are observed in both cultured cells and brain tissue.

It is also of interest to compare the ethanol tolerance and cross-tolerance seen in this cell culture system with the analogous changes seen in animal studies. For example, tolerance to the action of ethanol was detected at times as short as 5 min, and this raises the question of whether such rapid tolerance occurs to behavioral actions of ethanol in vivo or is peculiar to transfected cells. It is clear that a single injection of ethanol is sufficient to produce a marked tolerance (Crabbe et al., 1979; LeBlanc et al., 1975). However, few studies have determined the time course of this tolerance using behavioral measures. One of the studies that evaluated short times found that tolerance to ethanol induced ataxia developed within 3 to 5 min (Gill et al., 1993). Electrophysiological studies allow study of rapid tolerance in intact animals, and tolerance to the effects of ethanol on Purkinje cell firing in rat cerebellum can be seen within 5 min after application of ethanol (Pearson et al., 1996; Sorensen et al., 1980). As noted above, complete tolerance to ethanol enhancement of GABA_A receptor function can be seen in brain membranes from mice given a single injection of ethanol 5 min before death (Allan and Harris, 1987). Thus, the rapid development of tolerance to ethanol that occurs with transfected cells is consistent with behavioral, electrophysiological and biochemical results from animals treated with ethanol.

Another aspect of our results is that cross-tolerance occurred to flunitrazepam but not pentobarbital, and it is of interest to note that behavioral studies have noted acute cross-tolerance between ethanol and benzodiazepines but not between ethanol and pentobarbital (Khanna et al., 1992).

Studies of different types of cells demonstrated that at least some of the effects of chronic ethanol treatment are not critically dependent on subunit composition. For example, the decrease in muscimol E_max was observed with all subunit combinations tested, and the reduction in flunitrazepam action was equal in cells expressing gamma-2L and gamma-2S. These findings do not support the idea that the chronic effects of ethanol are related to the acute effects because only cells expressing the gamma-2L subunit are sensitive to ethanol concentrations of <100 mM (Harris et al., 1997). Studies of pentobarbital action show the selectivity of the adaptation produced by ethanol exposure. Pentobarbital produces two distinct actions on GABA_A receptors: an enhancement of GABA (or muscimol) action and a direct, GABA-independent, activation of the receptor. These effects appear to be due to different mechanisms (Sanna et al., 1995), and the direct effect is particularly pronounced for receptors containing the alpha-6 subunit (Thompson et al., 1996). Although chronic ethanol exposure reduced the action of four other modulators of GABA_A receptors of stably transfected cells, it did not alter the ability of pentobarbital to potentiate muscimol action, and it only slightly enhanced the direct action of pentobarbital. Thus, chronic ethanol treatment did not produce a general attenuation of GABA_A receptor function or reduction of all allosteric modulation of the receptor. In comparison to animal studies, the direct action of pentobarbital is attenuated by ethanol treatment of rats (Morrow et al., 1988), but pentobarbital enhancement of muscimol action is not changed by ethanol consumption in mice (Buck and Harris, 1990a, 1990b).

Given that our data suggest multiple mechanisms for receptor adaptation to ethanol in these cells, it is of interest to consider changes that could occur in these cells and, potentially, in animals. Because the cells contain defined subunits with expression controlled by the dexamethasone-sensitive promoter, it is unlikely that ethanol changes subunit expression. This is also indicated by a lack of change in surface receptor density in this study and subunit mRNA and protein levels noted in a previous study (Klein et al., 1995b). Thus, post-translational modifications seem the most likely candidates. Changes in subunit assembly, protein trafficking or subunit conformation are possible mechanisms. Also, GABA_A receptor subunits are phosphorylated by several protein kinases that alter the sensitivity of the receptor to agonist activation (Krishek et al., 1994; Moss and Smart, 1996; Valenzuela et al., 1995). In addition, the action of ethanol on the GABA_A receptor appears to depend on PKC activity (Harris et al., 1995a; Mihic and Harris, 1996; Weiner et al., 1997), and chronic ethanol treatment of cultured cells increases PKC activity (Messing et al., 1991; Roivainen et al., 1995). However, in the present study, an inhibitor of PKC failed to alter the reduction in muscimol action produced by chronic ethanol exposure. Given the evidence that there are multiple mechanisms for adaptation of GABA_A receptors to ethanol in these cells and that multiple kinases regulate these receptors, it is possible that other kinase inhibitors or activators might alter some of the responses to chronic ethanol treatment.

In summary, exposure of cultured cells to ethanol produces rapid and complex adaptation of GABA_A receptors stably transfected into these cells. Many of the changes in receptor function are remarkably similar to changes seen in animals given ethanol, suggesting that post-translational changes may be responsible for some of the changes in GABA_A receptor function produced by ethanol exposure in vitro and ethanol consumption in vivo.