Chronic ethanol consumption enhances internalization of α1 subunit-containing GABAA receptors in cerebral cortex

Authors


Address correspondence and reprint requests to A. Leslie Morrow, PhD, Center for Alcohol Studies, CB#7178, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599–7178, USA. E-mail: morrow@med.unc.edu

Abstract

The molecular mechanisms that underlie ethanol dependence involve alterations in the functional properties and subunit expression of GABAA receptors. Chronic ethanol exposure decreases GABAA receptor α1 subunits and increases α4 subunit levels in cerebral cortical membranes. This study explored the effect of chronic ethanol exposure on internalization of GABAA/benzodiazepine receptors. Chronic ethanol exposure increased α1 subunit levels by 46 ± 12% and [3H]flunitrazepam binding by 35 ± 9% in the clathrin-coated vesicle (CCV) fraction. There was a corresponding 34 ± 8% decrease in α1 peptide expression and 37 ± 6% decrease in [3H]flunitrazepam binding in the synaptic fraction. Chronic ethanol consumption also increased the α1 subunit immunoprecipitate in the cytosolic fraction (77 ± 22%), measured by western blot analysis. Moreover, co-immunoprecipitation of both clathrin and adaptin-α with α1 subunits was increased in the cytosolic fraction, suggesting that α1 subunit endocytosis is enhanced by chronic ethanol consumption. In contrast, α4 subunit peptide levels were not altered in the CCV fraction despite a 39 ± 13% increase in peptide levels in the synaptic fraction of cortex. Moreover, acute ethanol exposure did not alter α1 subunit peptide expression or [3H]flunitrazepam binding in the synaptic or CCV fractions. These results suggest that chronic ethanol consumption selectively increases internalization of α1 subunit-containing GABAA receptors in cerebral cortex.

Abbreviations used
AP-2

adaptor complex-2

CCV

clathrin-coated vesicles

GABAA

γ-aminobutyric acidA

PAGE

polyacrylamide gel electrophoresis

PBS

phosphate-buffered saline

PKC

protein kinase C

SDS

sodium dodecyl sulfate

GABAA receptors are ligand-gated ion channels that mediate inhibitory neurotransmission throughout the CNS. Molecular cloning has revealed a range of GABAA receptor subunits that can be divided by homology into subunit classes with multiple members: α(1–6), β(1–3), γ(1–3), δ, ε, θ and π. The majority of GABAA receptors are composed of two α, two β and one γ subunit (Im et al. 1995; Chang et al. 1996; Tretter et al. 1997; Farrar et al. 1999), where the γ subunit is located between an α and β subunit (Tretter et al. 1997; Sieghart et al. 1999) and expressed on cell membranes to mediate neuronal signaling (Sieghart 2000). Recombinant expression studies have clearly demonstrated that subunit composition determines the functional properties of receptor subtypes (Wisden and Seeburg 1992). Chronic ethanol consumption alters GABAA receptor function, producing cellular tolerance to GABA and ethanol, cross-tolerance to benzodiazepines and barbiturates, and sensitization to inverse agonists (Morrow et al. 1988; Tabakoff and Hoffman 1988; Buck and Harris 1990; Devaud et al. 1996, 1997; Devaud and Morrow 1998). Both sensitization to anti-convulsant actions and tolerance to anxiolytic effects of neurosteroids have been observed (Devaud et al. 1996; Kang et al. 1998). The mechanisms responsible for adaptation of GABAA receptors to chronic ethanol exposure are postulated to involve ethanol-induced changes in GABAA receptor subunit composition (Morrow 1995), or phosphorylation (Harris et al. 1998), yet conclusive evidence for these mechanisms is lacking (Grobin et al. 1998). Recent studies suggest that GABAA receptor function can be modulated by altered receptor trafficking (Brandon et al. 2002). Therefore, we sought to explore this possibility in ethanol-dependent rats.

Chronic ethanol exposure alters the expression of GABAA receptors in cerebral cortical membranes (see Grobin et al. 1998 for review). Alpha1 subunit peptide and mRNA expression are reduced (Morrow et al. 1990; Montpied et al. 1991; Mhatre and Ticku 1992; Devaud et al. 1995, 1997), and genetic deletion of this subunit produces reductions in GABA-mediated inhibition and increased CNS excitability similar to ethanol dependence (Kralic et al. 2002). In addition, there are increases in α4 subunit mRNA and peptide expression (Devaud et al. 1995, 1997; Mahmoudi et al. 1997) that have been linked to reductions in benzodiazepine sensitivity in both recombinant systems (Yang et al. 1995; Whittemore et al. 1996) and native receptors (Smith et al. 1998). The changes in membrane expression of GABAA receptor subunits following chronic ethanol exposure could involve altered receptor trafficking that determines cell surface receptor subtype expression and produces altered synaptic responses.

Receptors on the synaptic membrane may be regulated by trafficking via intracellular vesicles. Golgi-derived vesicles transport newly synthesized receptors to the cell surface, whereas endocytosis leads to degradation or recycling back to the cell surface. Selective endocytosis and/or recycling of receptors can alter receptor subunit expression or composition. For example, selective loss and endocytosis of γ2 subunit-containing GABAA receptors is observed following protein kinase C (PKC) activation in A239 cells (Connolly et al. 1999), while ethanol exposure increases peptide levels of PKCε and δ in PC12 cells (Messing et al. 1991). Furthermore, chronic lorazepam administration to mice causes increased sequestration of [3H]flunitrazepam binding sites and enhanced immunoreactivity of GABAA receptor α1 subunits in clathrin-coated vesicles (CCV) (Tehrani and Barnes 1997).

The present study explored the possibility that changes in GABAA receptor α1 and α4 subunit peptide expression in cortical membranes following chronic ethanol exposure are due to selective alterations in receptor internalization. Internalization was assessed by determination of peptide levels in cytosolic and CCV versus synaptic membrane fractions of cerebral cortex by western blot analysis, and quantified by measuring [3H]flunitrazepam binding. Furthermore, we investigated whether adaptor complex-2 (AP-2) and clathrin are involved in the internalization of GABAA receptor α1 subunits following chronic ethanol exposure.

Materials and methods

Dynabeads, Protein A and magnets were purchased from Dynal Inc. (Lake Success, NY, USA). Rabbit polyclonal α1 and α4 antibodies were obtained from Werner Sieghart and guinea pig α1 antibody was obtained from Jean-Marc Fritschy. Clathrin heavy chain and adaptin-α antibodies were purchased from BD Transduction Laboratory (Lexington, KY, USA). Electrophoresis was performed using 8–16% Tris-Glycine gels from Novex (San Diego, CA, USA). [3H]Flunitrazepam (71 Ci/mm) was purchased from NEN (Boston, MA, USA) and diazepam was obtained from Biomol Research Laboratory (Plymouth Meeting, PA, USA).

Chronic and acute ethanol administration

All studies were carried out in accordance with the Guide for the Care and Use of Laboratory Animals and were approved by the Institution of Animal Care and Use Committee at the University of North Carolina. Male Sprague-Dawley rats were housed individually and administered a nutritionally-complete liquid diet. Ethanol (6% v/v) was given for one week followed by 7.5% (v/v) ethanol for the second week; control animals were pair-fed the identical diet with dextrose substituted for ethanol (Devaud et al. 1997). Water was available ad libitum. On day 15, animals were killed and cerebral cortex was dissected over ice and stored at −80°C. For acute ethanol exposure, Sprague-Dawley rats received an ethanol injection (2 g/kg, i.p.) 1 h prior to decapitation and collection of cerebral cortex.

Tissue and protein preparations

Cytosolic fraction from individual cerebral cortices was collected by homogenization, low-speed centrifugation in 0.32 m sucrose, and centrifugation of supernatant fluid at 12 000 g for 20 min followed by 100 000 g centrifugation for 30 min. The supernatant fluid (cytosolic fraction) was collected and stored at − 80°C after adding protease inhibitors (leupeptin, 1 mg/mL; pepstatin A, 0.7 mg/mL; trypsin inhibitors, 0.1 mg/mL; and phenylmethylsulfonyl fluoride, 1 mm).

The CCV and synaptic membrane fractions were prepared from cerebral cortex as described by Tehrani and Barnes (1997) with minor modifications. Briefly, bilateral cerebral cortices from eight animals per experimental group were pooled and homogenized in 3 volumes of isolation buffer [10 mm 2-(n-morpholino)ethanesulfonic acid, pH 6.5, 100 mm KCl, 1 mm EGTA, 0.5 mm MgCl2, 0.3 mm phenylmethylsulfonyl fluoride (PMSF), 50 µm dithiothreitol DTT], 2 mm benzamidine, 0.1 mg/mL bactracin and 0.1 mg/mL soyabin trypsin inhibitor]. The homogenate was centrifuged at 20 000 g for 20 min. The supernatant fluid was collected and centrifuged at 100 000 g for 1 h to obtain a microsomal pellet. This pellet was resuspended in 2 mL isolation buffer, mixed with 2 mL of a solution containing 12.5% ficoll 400 and 12.5% sucrose in isolation buffer, and then centrifuged at 42 000 g for 40 min. The resulting supernatant fluid was collected and mixed with 3 volumes of isolation buffer containing Triton X (0.1%). The mixture was incubated for 30 min on ice and then centrifuged at 100 000 g for 90 min. The resulting pellet (CCV) was resuspended in isolation buffer and stored at − 80°C. To obtain the synaptic fraction, the pellet resulting from the 20 000 g centrifugation described above was washed three times with Tris buffer at 45 000 g for 30 min, resuspended in isolation buffer and stored at − 80°C.

Immunoprecipitation

GABAA receptors in the cytosolic fraction were immunoprecipitated with α1 subunit-specific (rabbit) antibody as previously described (Kumar et al. 2002), except that the ratio of antibody/receptor was increased in order to detect changes in cytosolic receptor levels. Briefly, cytosolic protein (425 µg) was solubilized using a solution containing 1% (w/v) Triton X-100, 1% (w/v) sodium deoxycholate, 0.1% (w/v) sodium dodecyl sulfate (SDS). The α1 subunit-specific antibody was linked to magnetized Dynabeads by incubating 125 µL Dynabeads with 12.5 µL α1 subunit-specific antibody (628 µg/µL) for 1 h. The prepared beads were mixed with solubilized receptors in a final volume of 500 µL and incubated overnight at 4°C. The receptor-antibody-beads were then washed three times with phosphate-buffered saline (PBS). After the final wash, the receptor-antibody-bead complex was resuspended in 50 µL SDS and boiled for 5 min. Beads were separated from the immunoprecipitate by exposure to a magnet for 2 min, followed by collection of the supernatant fluid and denaturation with 1 µL 5 m DTT.

Western blot analysis

The immunoprecipitate and various subcellular fractions were analyzed by western blot analysis under conditions of protein linearity (Devaud et al. 1997). Briefly, subcellular fractions and α1 immunoprecipitate were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) using Novex Tris-Glycine gels (8–16%), transferred to polyvinylidene diflouride membrane (Invitrogen, Carlsbad, CA, USA) and immunoblotted with GABAA receptor α1 (guinea pig) or α4 subunit (rabbit) antibodies as described by Kumar et al. (2002). Bands were detected by enhanced chemiluminescence (Amersham, Rockford, IL, USA), apposed to X-ray films under non-saturating conditions, and analyzed by densitometric measurements using NIH Image 1.57. All comparisons were made within blots. Blots were subsequently exposed to a second primary antibody, directed against β-actin, to verify equivalent protein loading. Statistical analysis was conducted using the Student's t-test.

Radioligand binding

[3H]Flunitrazepam binding was performed according to Tehrani and Barnes (1997), with some modifications. Briefly, synaptic membrane fraction (100 µg) or CCV fraction (100 µg) was added to assay buffer containing 50 mm Tris-HCl, pH 7.4, 120 mm NaCl, 5 mm KCl and 10 nm[3H]flunitrazepam in a final volume of 500 µL. Diazepam (10 µm) was used to define non-specific binding. Samples were incubated on ice for 90 min. The reaction was terminated by rapid filtration under vacuum (0.84 atm Hg) using Whatman GF/B filter strips. Samples were washed twice with 3 mL aliquots of assay buffer at 0 − 4°C. Filters were dried, added to liquid scintillation cocktail and counted by liquid scintillation spectroscopy.

Results

Chronic ethanol exposure to rats increased α1 subunit peptide levels by 46 ± 12% (p < 0.01, n = 4) in the CCV fraction of cerebral cortical membranes (Fig. 1a). In the synaptic membrane fraction of cortex from the same rats, GABAA receptor α1 subunit peptide levels were decreased by 34 ± 8% (p < 0.01, n = 4) (Fig. 1b). Peptide measurements were conducted using 35 µg CCV and 25 µg synaptic membrane proteins, determined to be in the linear range of α1 subunit peptide detection (Figs 1c and d). These results show that internalization of α1 subunits accompanies the decreased expression of α1 subunits in the synaptic fraction.

Figure 1.

Chronic ethanol consumption increases GABAA receptor α1 subunit peptides in the CCV fraction and decreases α1 subunit peptides in the synaptic fraction. Representative western blot analysis showing α1 subunit peptides in (a) the CCV fraction and (b) the synaptic fraction of cerebral cortex from control (lane 1) and ethanol-dependent (lane 2) rats. Each independent experiment was conducted in duplicate using CCV and synaptic fractions prepared from pooled cerebral cortices of 8 rats/group. Alpha1 subunit labeling was normalized to β-actin to control for equivalent gel loading and transfer. Chronic ethanol consumption increased α1 subunit peptide level in the CCV fraction expression by 46 ± 12% (**p < 0.01, n = 4). In contrast, α1 subunit expression in the synaptic fraction (b) was decreased by 34 ± 8% (p < 0.01, n = 4). Detection of α1 subunit peptide expression was linear with protein concentration in the (c) CCV fraction (5–80 μg) and (d) synaptic fraction (20–100 μg).

The effects of chronic ethanol administration on GABAA receptor α1 subunit peptide levels were not observed following acute ethanol administration. Acute ethanol administration (2.0 g/kg) did not alter α1 subunit peptide levels in either the CCV fraction (Fig. 2a) or the synaptic membrane fraction (Fig. 2b). Hence, acute administration of intoxicating ethanol concentrations does not alter internalization of α1 subunits.

Figure 2.

Acute ethanol exposure does not alter GABAA receptor α1 subunit peptide expression in the CCV fraction or the synaptic fraction of cerebral cortex. Rats were injected with ethanol (2 g/kg, i.p.) or saline and cerebral cortices were collected following decapitation after 1 h. The CCV and synaptic fractions were isolated, analyzed by western blot analysis and probed using various antibodies. Representative western blot analysis showing α1 subunit peptides and β-actin in (a) the CCV fraction and (b) the synaptic fraction of cerebral cortex from rats administered saline (lane 1) and ethanol (lane 2). Alpha1 subunit labeling was normalized to β-actin to control for equal gel loading and transfer. Acute ethanol exposure did not alter α1 subunit peptide expression in the (a) CCV (n = 4) or (b) synaptic fraction (n = 4) of cerebral cortex. Independent experiments were conducted in duplicate using CCV and synaptic fractions from pooled cerebral cortices of 8 rats/group.

In order to determine whether chronic ethanol consumption alters the subcellular localization of benzodiazepine receptors, we performed [3H]flunitrazepam binding in both the synaptic membrane and CCV fractions. [3H]Flunitrazepam binds to both Type I and Type II benzodiazepine binding sites containing α1, α2, α3 and α5 subunits. As cerebral cortex predominantly contains Type I receptors that contain α1 subunits (Kralic et al. 2002), any change in [3H]flunitrazepam binding would largely reflect changes in α1 subunit-containing receptors. Chronic ethanol consumption decreased [3H]flunitrazepam binding in the synaptic fraction by 37 ± 6% (p < 0.0001, n = 4) and increased [3H]flunitrazepam binding in CCV preparations by 35 ± 9% (p < 0.05, n = 4) (Table 1). In contrast, acute ethanol administration did not alter [3H]flunitrazepam binding in either synaptic membrane or CCV fractions (Table 1).

Table 1.  [3H]Flunitrazepam (10 nm) binding activity in CCV and synaptic membrane fraction of cerebral cortex
Experimental groupCCV fraction specific binding (fmol/mg)Percent changeSynaptic fraction specific binding (fmol/mg)Percent change
  • The values are mean ± SEM from four membrane preparations performed in triplicate.

  • *

    p < 0.05 as compared with control;

  • ***

    p < 0.0001 as compared with control.

Control (pair-fed)159 ± 1.5 822.9 ± 45.6 
Chronic ethanol216 ± 15.7*35 ± 9%514.8 ± 29.9***−37 ± 6%
Control (i.p. saline injection)151 ± 5.46 866.6 ± 61.7 
Acute ethanol (i.p. ethanol injection)161.3 ± 0.66 ± 3.3%873.4 ± 40.70.8 ± 8%

To determine whether chronic ethanol consumption produced a generalized increase in receptor internalization, we measured GABAA receptor α4 subunit peptides in both CCV and synaptic membrane fractions. We found that α4 subunit peptide levels were increased in the synaptic fraction by 39 ± 13% (p < 0.05, n = 4) following chronic ethanol exposure (Fig. 3a), with no change in α4 subunit peptide levels in the CCV fraction (Fig. 3b). Thus, chronic ethanol administration has selective effects on the internalization of α1 but not α4 subunit-containing GABAA receptors.

Figure 3.

Chronic ethanol administration increases α4 subunit peptide expression in the synaptic fraction of cerebral cortex, with no change in the CCV fraction. CCV and synaptic fractions were purified and analyzed by western blot analysis. Representative western blot analysis showing α4 subunit peptides in (a) the CCV fraction and (b) the synaptic fraction of cerebral cortex from control (lane 1) and ethanol-dependent (lane 2) rats. Alpha4 subunit level was normalized to β-actin. Chronic ethanol consumption did not alter α4 subunit peptide level in the CCV fraction. In contrast, α4 subunit expression in the synaptic fraction (b) was increased by 39 ± 13% (p < 0.05, n = 4). Independent experiments were conducted in duplicate using CCV and synaptic fractions from pooled cerebral cortices of 8 rats/group.

To determine whether internalization of α1 subunit-containing GABAA receptors involves direct binding to adaptor complex-2 and clathrin heavy chain, GABAA receptors in the cytosolic fraction of cerebral cortex were solubilized and immunoprecipitated with α1 subunit antibody. The immunoprecipitate was analyzed by western blot analysis and probed with various antibodies. We detected a 77 ± 22% increase (p < 0.05, n = 4 pairs) in α1 subunit peptide levels in the α1 immunoprecipitate (Fig. 4a). This increase was probably due to increased levels of intracellular α1 subunit receptors, as shown above. In addition,co-immunoprecipitation of adaptin-α and clathrin heavy chain with α1-containing GABAA receptors demonstrated an association of these proteins in the cytosolic fraction. Furthermore, the association of both adaptin-α and clathrin with the α1 subunit immunoprecipitate was increased following chronic ethanol exposure (Figs 4c and d). Co-immunoprecipitation of adaptin-α was increased 213 ± 43% (p < 0.01, n = 4) and association with clathrin heavy chain was increased 192 ± 45% (p < 0.01, n = 4) following chronic ethanol exposure. Since these dramatic increases in co-immunoprecipitation could be partly due to increased immunoprecipitation of α1 subunits, the level of adaptin-α and clathrin heavy chain were normalized to the level of α1 subunit peptide in the corresponding immunoprecipitates. Upon normalization, co-immunoprecipitation of both adaptin-α (87 ± 33%, p < 0.05, n = 4) and clathrin (62 ± 17%, p < 0.05, n = 4) with GABAA receptor α1 subunits was increased by chronic ethanol consumption. Therefore, association of clathrin and adaptin-α with α1 subunit-containing GABAA receptors is likely to mediate the enhancement of the internalization of α1 subunit-containing GABAA receptors following chronic ethanol exposure.

Figure 4.

Chronic ethanol exposure increases α1 subunit peptide associated with adaptin-α and clathrin in the cytosolic fraction of cerebral cortex. Cytosolic fraction of cortex was solubilized and immunoprecipitated with rabbit α1 subunit antibody. The immunoprecipitate was subjected to SDS-PAGE and probed with (a) α1 subunit-specific antibody, (b) adaptin-α and (c) clathrin heavy chain antibodies. Lane 1 (control) and lane 2 (chronic ethanol-treated) cortex are α1 immunoprecipitates. (a) Alpha1 peptide level was increased by 77 ± 22% (*p < 0.05, n = 4 pairs) in the cytosolic fraction following chronic ethanol exposure. In addition, co-immunoprecipitation of adaptin-α (b) was increased by 213 ± 43% (**p < 0.01, n = 4 pairs), and co-immunoprecipitation of clathrin (c) was increased by 192 ± 45% (p < 0.01, n = 4 pairs), following chronic ethanol consumption compared with pair-fed controls. The levels of adaptin-α (b) and clathrin heavy chain (c) were normalized to the level of α1 subunit peptide in corresponding immunoprecipitate. Upon normalization, co-immunoprecipitation of both adaptin-α (87 ± 33%, p < 0.05) and clathrin (62 ± 17%, p < 0.05) with α1 subunits were increased following chronic ethanol consumption.

Discussion

The present study demonstrates that chronic ethanol consumption selectively alters the subcellular localization of α1 subunit-containing GABAA receptors in cerebral cortex of rat. Internalization of α1 subunit-containing receptors is increased, since there was a simultaneous increase in both α1 subunit peptide and [3H]flunitrazepam binding in the CCV fraction and a comparable decrease in these measures in the synaptic fraction of cerebral cortex. In contrast, internalization of α4 subunit-containing receptors is unchanged following chronic ethanol exposure. Furthermore, acute ethanol administration does not alter internalization of α1 subunit-containing receptors, as α1 subunit peptide expression in the synaptic fraction or CCV fraction was not altered. Moreover, there was no change in [3H]flunitrazepam binding in the CCV fraction or the synaptic fraction following acute ethanol exposure. Sequestration into CCVs is one method of receptor down-regulation. Thus, these data suggest that altered expression of α1 subunits at the cell surface following chronic ethanol administration may be due to selective alteration in the trafficking of GABAA receptor α1 subunit-containing receptors.

The ability of chronic ethanol exposure to increase the number of internalized α1 subunit GABAA receptors could be due to increased endocytosis, decreased recycling and/or decreased degradation. Since chronic ethanol administration increased the co-immunoprecipitation of the endocytotic proteins clathrin and adaptin-α, it is likely that the increased level of internalized receptors was the result of increased endocytosis of these receptors via association with adaptin-α and clathrin. However, the possibility that ethanol also alters receptor recycling or degradation cannot be excluded by the present data.

The observation that acute ethanol exposure did not alter α1 subunit expression and [3Η]flunitrazepam binding in CCV or synaptic fractions suggests that acute ethanol exposure does not alter endocytosis of GABAAα1 subunit receptors. However, previous studies have shown that GABAA receptors are internalized within minutes following PKC activation in vitro (Chapell et al. 1998; Connolly et al. 1999). It is possible that the GABAA receptor α1 subunits are endocytosed following acute ethanol exposure, but receptor recycling compensates and maintains the receptor expression detected in the synaptic membrane and CCV fractions. Following chronic ethanol exposure, this mechanism may be insufficient to compensate for the apparent increase in the endocytosis of the receptors. This could account for the decrease in [3Η]flunitrazepam binding on the cell surface and the subsequent decrease in synaptic efficacy that has been observed. Previous studies have shown that chronic ethanol exposure does not result in decreased [3H]flunitrazepam binding in membrane homogenates of cerebral cortex (Karobath et al. 1980; Rastogi et al. 1986; Mehta and Ticku 1999). However, these studies did not distinguish between synaptic membrane and intracellular receptors. Hence, it is not surprising that differences were not found, since we show evidence for alterations in the subcellular localization (not the total number) of receptors.

It is possible that chronic ethanol exposure non-specifically increases the endocytotic process in cortical cells that may lead to increased internalization of α1 subunit-containing GABAA receptors. However, we did not detect an increase in GABAA receptor α4 subunit in the CCV fraction following chronic ethanol exposure. Furthermore, there was no alteration of clathrin heavy chain immunoreactivity in the CCV fraction following chronic ethanol consumption (data not shown). Hence, ethanol-induced changes in subunit expression may involve altered endocytosis of specific receptor subtypes.

Endocytosis via CCVs requires association of adaptor complex-2 (AP-2) and clathrin. AP-2 is specific for endocytosis of surface proteins and consists of four subunits: β, µ, α and σ (Hirst and Robinson 1998). Adaptin-α is specific for AP-2, and it binds to the cytosol-facing domain of plasma membrane receptor proteins that become internalized via clathrin-coated pits but does not bind to plasma membrane proteins that are not internalized via clathrin-coated pits (Lodish et al. 1996). Therefore, if AP-2 and clathrin are attached to a receptor, it is either in the process of internalization or already internalized. We detected adaptin-α and clathrin associated with α1 subunit-containing GABAA receptors in the cytosolic fraction of both control and ethanol-exposed rats, indicating that α1 subunit-containing GABAA receptors are internalized under control conditions in rat cerebral cortex. Previous studies have also demonstrated a direct association of these proteins with GABAA receptors in vitro (Kittler et al. 2000). Furthermore, association of adaptin-α and clathrin with α1 subunit-containing GABAA receptors was increased following chronic ethanol consumption, probably due to increased endocytosis of these receptors.

Selective endocytosis of GABAA receptors and recycling back to the cell surface is one possible mechanism that may control the cell surface composition of GABAA receptors (Connolly et al. 1999). Receptor trafficking is important for receptor function. For example, blockade of constitutive endocytosis in hippocampal neurons leads to an increase in GABAA receptor-mediated current in these cells, indicating that turnover of these receptors contributes to GABA-mediated inhibition (Kittler et al. 2000). Thus, receptor trafficking can play an important role in GABAA receptor adaptations following chronic ethanol exposure.

Trafficking of receptors is a highly regulated mechanism. The cell signaling pathways that control endocytosis and recycling have not been fully elucidated. However, it is apparent that PKC activity plays a role. Phosphorylation by PKC may alter receptor subunit composition at the membrane surface via changes in receptor trafficking. Furthermore, chronic ethanol consumption results in decreased association of PKCγ with α1 subunit-containing GABAA receptors and decreased expression of α1 subunit at the cell surface. In contrast, chronic ethanol exposure results in increased association of PKCγ with α4 subunit-containing GABAA receptors and increased expression of α4 at the cell surface (Kumar et al. 2002). Moreover, GABAA receptor α4 subunits contain a consensus site for PKC phosphorylation (Wisden et al. 1991; Macdonald 1995; Mohler et al. 1996) that may influence receptor recycling. For example, threonine phosphorylation diverts internalized epidermal growth factor receptors from degradative pathway to recycling pathway (Bao et al. 2000). Thus, differential effects of ethanol on the association of PKCγ with these receptors and phosphorylation may influence endocytosis and/or recycling of these receptors in vivo. Previous studies in Xenopus oocyte and A293 cells suggest that PKC activation leads to increased internalization of α1βγ GABAA receptors due to reduced recycling of receptors (Chapell et al. 1998; Connolly et al. 1999). However, it is not known whether PKC activation produces internalization of any GABAA receptor in cerebral cortical neurons. The expression of PKC isoforms in brain differ from non-neuronal cells, and various isoforms of PKC have diverse effects on trafficking (Goode et al. 1994). In addition, it is likely that other proteins phosphorylated by PKC may regulate receptor trafficking. Proteins that may be important for GABAA receptor anchoring and/or trafficking could include gephyrin (Kneussel et al. 1999), AP-2 (Kittler et al. 2000) and/or GABARAP (Wang et al. 1999; Wang and Olsen 2000). Hence, PKC phosphorylation of GABAA receptor-associated proteins may alter GABAA receptor trafficking and subunit expression/composition on the cell surface following chronic ethanol exposure.

Chronic ethanol consumption results in ethanol dependence and alters properties of GABAA receptors in brain (Grobin et al. 1998; Cagetti et al. 2003). Ethanol tolerance and dependence are associated with a decrease in GABAA receptor-mediated responses in cerebral cortex (Morrow et al. 1988; Sanna et al. 1993). In cerebral cortex, muscimol-stimulated chloride flux is decreased following chronic ethanol inhalation (Morrow et al. 1988) or liquid diet administration (Devaud et al. 1996). Furthermore, there is decreased sensitivity to benzodiazepines and increased sensitivity to inverse agonists (Buck and Harris 1990). The selective loss of α1 subunit-containing GABAA receptors from the membrane surface following chronic ethanol exposure may explain decreased efficacy of GABA agonists since cerebral cortex predominantly contains α1 subunit-containing receptors. Furthermore, increased expression of α4-containing GABAA receptors following chronic ethanol exposure may contribute to reduced GABA and benzodiazepine responses, since recombinant GABAA receptors with α4β2γ2 subunits exhibit lower efficacy in response to GABA and benzodiazepine agonists than α1β2γ2 receptors (Whittemore et al. 1996). However, other mechanisms may also contribute to adaptations following chronic ethanol exposure. For example, post-translational modification of receptor subunits or altered anchoring of GABAA receptors may determine the intracellular localization and/or sensitivity of receptors. Further studies are needed to investigate these possibilities.

Ethanol-induced alterations in GABAA receptor subunit expression differ across brain regions. Chronic ethanol consumption decreases GABAA receptor α1 subunit peptide expression in cerebral cortex, but not hippocampus (Matthews et al. 1998) or hypothalamus (Devaud et al. 1999). Furthermore, there are regional differences across brain on the effects of ethanol on PKC expression (Pandey et al. 1993; Narita et al. 2001) and the association of PKCγ with α1 subunit-containing GABAA receptors (Kumar et al. 2002). Therefore, it is possible that PKC-mediated alterations of GABAA receptor trafficking may be responsible for regional differences in GABAA receptor subunit expression following chronic ethanol exposure. Hence, further studies are needed to investigate the effects of ethanol on receptor trafficking to obtain a clear understanding of GABAA receptor adaptations following chronic ethanol consumption.

Acknowledgements

The authors thank Jean-Marc Fritschy (Institute of Pharmacology and Toxicology, University of Zurich, Switzerland) and Werner Sieghart (Brain Research Institute, University of Vienna, Austria) for generously providing antibodies. This work was supported by National Institute of Health grants AA09013 and AA11605 (ALM).

Ancillary