Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Alcoholics undergo a functional adaptation of the central nervous system (CNS) that produces long-lived behavioral responses, such as tolerance, dependence, sensitization, and addiction. We have previously shown that chronic exposure to ethanol causes changes in gene expression in neural cells (Gayer et al., 1991; Miles et al., 1991, 1992, 1993, 1994). An altered expression of specific genes could underlie aspects of CNS adaptive behavioral responses seen with long-term exposure to ethanol. Similar adaptive changes in gene expression have been suggested to occur with other drugs of abuse (Nestler et al., 1993).

Previously, we found that the molecular chaperones Hsc70, GRP78, and GRP94 are a subset of ethanol-responsive genes in NG108-15 neuroblastoma cells (Miles et al., 1991, 1994; Hsieh et al., 1996). Hsc70 is also induced by ethanol in developing rat brain (Holownia et al., 1995). Because chronic ethanol exposure decreases receptor recycling and protein secretion (Casey et al., 1989; Henderson et al., 1989; Baskin, 1990; Casey et al., 1990; Dalke et al., 1990; Ghosh et al., 1991; Tuma et al., 1991), increased chaperone expression could represent a specific adaptive response to this action of ethanol. Molecular chaperones are also induced during other forms of CNS plasticity, perhaps functioning in synaptic remodeling (Kennedy et al., 1992).

The molecular chaperone Hsc70 is a constitutively expressed member of the 70-kDa stress protein gene family. The highly related protein Hsp70 is a major inducible gene product responding to heat shock and other stressors. Hsc70 is involved in several aspects of normal cellular protein trafficking, including folding of nascent polypeptide chains on polyribosomes (Beckmann et al., 1990), uncoating of clathrin-coated vesicles (Chappell et al., 1986), and transport of proteins into cellular organelles (Deshaies et al., 1988). Decreased expression of the yeast Hsc70 homolog produces corresponding decreases in protein secretion (Deshaies et al., 1988), suggesting a dynamic relationship between Hsc70 expression and protein trafficking.

Ethanol regulates Hsc70 abundance by increasing the rate of hsc70 transcription (Miles et al., 1991). The 2500 base pairs (bp) of DNA 5' to the hsc70 transcription start site contains cis-acting sequences that confer ethanol responsiveness on a reporter gene (Miles et al., 1991). Our previous studies suggested that ethanol, at concentrations seen in actively drinking alcoholics (Urso et al., 1981), induces hsc70 transcription through a mechanism different from a typical heat-shock response because ethanol does not induce the related hsp70 gene. Conversely, heat shock induces large increases in expression of Hsp70 but not Hsc70 in NG108-15 cells (Miles et al., 1991).

Identification of the cis-acting sequences and cognate DNA-binding proteins mediating ethanol responsiveness of the hsc70 gene could have implications in the regulation of other ethanol-responsive genes. If ethanol-induced changes in gene expression indeed contribute to CNS behavioral responses to ethanol, then a determination of how ethanol regulates gene transcription could produce novel strategies for intervention in alcoholism. We therefore performed a detailed study on ethanol regulation of hsc70 gene transcription. We found that ethanol induction of hsc70 transcription required a binding site for Sp1, a widely used transcriptional activator. These findings provide insight into the molecular basis of ethanol-induced changes in specific gene expression.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cell Culture and Transfection. NG108-15 cells were cultured in 10% Nuserum-containing medium as described previously (Miles et al., 1991). Transient transfections of NG108-15 cells were performed by electroporation using log phase cells at 6 × 10⁶ cells/ml in Dulbecco's modified essential medium. Cells were pulsed at 200 V and 1180 µF capacitance at room temperature, allowed to recover for 10 min, and then transferred to serum-containing medium in 6-well plates. Twenty-four hours later, cells were incubated in medium with or without ethanol at concentrations listed in the text. The next day, cells were harvested and assayed for chloramphenicol acetyl-transferase (CAT) activity as described previously (Miles et al., 1991). A single electroporation sample was used to seed paired control and ethanol-treated cells; this eliminated the need for internal transfection controls for comparing ethanol with control. However, in some cases, a -galactosidase reporter gene coupled to a cytomegalovirus long terminal repeat (LTR) was used as an internal control for transfection efficiency. In some experiments, LipofectAMINE (Life Technologies, Grand Island, NY) was used to transfect NG108-15 cells with results similar to those seen with electroporation. Drosophila embryonic Schneider line 2 (SL2) cells were cultured in Schneider's Drosophila medium (Life Technologies) supplemented with 10% heat-inactivated fetal calf serum at 25°C without CO₂ in 25-cm² flasks (Corning). SL2 cells were grown at a density of 1-5 × 10⁶/ml and split into fresh medium every 3 days. For cotransfection studies, 60-mm dishes were seeded at 1.5 × 10⁶ cells/well 24 h before transfection by calcium phosphate precipitation (Kingston, 1987). Amounts of reporter plasmid and Sp1 expression plasmid used for transfections are as reported in the legend to Fig. 5. Transfected cells were harvested after 48 h and assayed for CAT activity.

Plasmid Constructions. A plasmid containing the 2500 to +60-bp region of the rat hsc70 gene coupled to a CAT reporter gene (pHsc2500) was the generous gift of Dr. Hugh Pelham (Cambridge, UK). Progressive deletions of the hsc70 promoter were made by digestion with appropriate restriction endonucleases (Boehringer Mannheim, Indianapolis, IN) or with exonuclease III and mung bean nuclease (Stratagene, La Jolla, CA). Deletions and point mutations in the hsc70 promoter sequences proximal to 113 (see Fig. 3) were made using homologous or mutated oligonucleotides (Table 1) as primers in polymerase chain reaction (PCR) amplifications (pHsc37, pHsc74, pHsc74m62). The 3' end of these fragments extended to the +10 position of the hsc70 5'-untranslated region. PCR products were inserted into a CAT vector constructed by deletion of all hsc70 promoter sequences from pHsc2500. The composition of each plasmid construct described was confirmed by DNA sequence analysis using the dideoxynucleotide-chain termination technique (Sanger et al., 1977).

Primer Extension Analysis. Total RNA was isolated (Miles et al., 1994) from NG108-15 cells transiently transfected as described in Results. RNA (10 µg) was then used for primer extension analysis as described previously (Thibault et al., 1999). An antisense oligonucleotide, complementary to the CAT coding region sequence ⁺14 to ⁺33 (5'-AACGGTGGTATATCCAGTGA-3') was end-labeled with [-³²P]ATP (DuPont-New England Nuclear, Boston, MA) by polynucleotide kinase and used for primer extension. Products were analyzed on a 6% denaturing polyacrylamide gel next to size markers. RNA from nontransfected cells was used as a control for nonspecific extension products.

DNA-Binding Protein Analysis. Whole-cell lysates were made from NG108-15 cells treated with or without ethanol as described in the text. Lysates were made (Zimarino and Wu, 1987) in buffer containing 10 mM HEPES, pH 7.9, 25% glycerol, 0.42 M KCl, 1.5 mM MgCl₂, 0.2 mM EDTA, 0.5 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride, and 2 µg/ml concentration each of the protease inhibitors aprotinin, leupeptin, and soybean trypsin inhibitor and were stored in aliquots at 85°C. Purified Sp1 was obtained from Promega. Antisera against Sp1 or Sp3 were obtained from Santa Cruz Biotechnology (Santa Cruz, CA).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Deletion Analysis of hsc70 Promoter Identifies a cis-Acting Region Required for Ethanol Responsiveness. The 2500 to +60 region of the rat hsc70 gene contains cis-acting elements that confer ethanol responsiveness in stable transfection analyses (Miles et al., 1991). This region contains several known promoter motifs (Fig. 1A), including two stretches of consensus heat shock elements (HSEs; Sorger and Pelham, 1987). We first verified that transient transfection studies would show an ethanol response similar to that seen on stable transfection studies (Miles et al., 1991). Plasmid pHsc2500 showed an ethanol concentration-dependent increase in CAT expression that plateaued at nearly 2-fold control levels with 100 mM ethanol (Fig. 2A). In contrast, a control plasmid containing the MSV LTR (pMSV) had no response to ethanol. The magnitude and concentration response of the Hsc70 induction in these transient assays closely paralleled results from previous stable transfection studies (Miles et al., 1991). Furthermore, the concentration response resembles that previously determined for induction of Hsc70 mRNA and protein (Miles et al., 1991).

View larger version (29K): [in this window] [in a new window]   Fig. 1.   Organization of the hsc70 promoter and identification of conserved elements. The organization of the rat hsc70 promoter is diagrammed (A) based on prior work by Sorger and Pelham (1987). The position of putative Sp1-binding sites and HSEs are indicated. Alignment of hsc70 promoter regions from rat, mouse, hamster, and human (B) identified an area of extensive sequence conservation within the proximal 70 bp of the rat promoter. GenBank identification numbers for sequences are indicated. Alignment of available sequence for more distant regions showed that significant homology existed only in the area of the distal HSE indicated in A (not shown).

View larger version (23K): [in this window] [in a new window]   Fig. 2.   Transient transfection analysis of hsc70 promoter: Proximal 74 bp region responds to ethanol (EtOH) but not arsenite (Ars). A, NG108-15 cells were transfected by electroporation with plasmids containing 2500 bp of the hsc70 promoter (pHsc2500) or the LTR region of MSV (pMSV). Twenty-four hours after transfection, cells were treated with the indicated concentrations of ethanol. After an additional 24-h incubation, cells were harvested and assayed for CAT activity. Results are expressed as a percent of CAT activity of control (mock-treated) cells transfected in the same cuvette as ethanol-treated cells. Assays were performed on triplicate wells of cells in each experiment. Results are the mean ± S.E. from experiments repeated 4 (50 and 100 mM) or 33 (200 mM) times. The pHsc2500 data showed significant differences with ethanol treatment (*P < .05, **P < .005, ***P < .0001; single-group t test with Bonferroni correction for multiple groups). B, NG108-15 cells were transfected as above with pHsc2500 or pHsc74. Twenty-four hours after transfection, cells were treated with saline, ethanol (200 mM), or arsenite (25 µM) for 6 h followed by lysis and assay for CAT activity. Results show absolute CAT activity and are the mean ± S.D. for triplicate wells of cells. Results are representative of experiments repeated three times. Con, control.

View larger version (40K): [in this window] [in a new window]   Fig. 3.   Identification of ethanol (EtOH) responsive cis-acting sequences in hsc70 promoter. Deletion or point mutations in the hsc70 promoter were made as described in Materials and Methods. Transfections were performed in triplicate as described in Fig. 2. After transfection, cells were treated for 24 h with or without ethanol (100 mM) followed by lysis and assay for CAT activity. A, results from a representative experiment are expressed as absolute CAT activity. Error bars indicate S.D. B, results are presented as the percentage of activity in control (Con) cells. Percent control results are the mean ± S.E. from experiments repeated at least six times. Basal activity results represent the fold above background (mock-transfected) CAT activity. Statistical significance was determined by single-group t test analysis with Bonferroni correction for multiple groups (for differences from control cells) or ANOVA with Scheffé's F test post hoc analysis to determine differences between various constructs: *P < .05 versus control (t test); P < .05 versus pHsc74 (ANOVA, Scheffé's post hoc).

View larger version (34K): [in this window] [in a new window]   Fig. 4.   Primer extension analysis of transient transfection products. Total RNA prepared from cells transfected with the indicated plasmids was analyzed by primer extension. Arrows indicated expected transcript sizes for pHsc constructs (~160 bp) or pMSV (~110 bp). Ethanol treatment caused a significant increase in extended product for cells transfected with pHsc2500, pHsc241, or pHsc74. C, control; E, ethanol.

Point Mutation Analysis Shows that an Sp1 Consensus Binding Site Is Required for Ethanol Responsiveness. A point mutation in the Sp1 site (pHsc74 m62) caused a large decrease (P < .05, ANOVA) in ethanol responsiveness compared with the parent construct pHsc74 (Fig. 3, A and B). Construct pHsc74 m62 was identical with pHsc74 except for a single base mutation that changed the Sp1 consensus site (67 to 61) from GGGGCGG to GGGGCTG. Although pHsc74 m62 still showed some residual induction by ethanol (148 ± 23% of control), this did not reach statistical significance (P > .08). All of the hsc70 deletion and point mutations constructs had basal CAT activity at least 2-fold above background levels obtained from mock-transfected cells (Fig. 3B).

View larger version (25K): [in this window] [in a new window]   Fig. 5.   Response of hsc70 promoter constructs to Sp1 expression in Drosophila SL2 cells. Schneider line 2 cells (SL2 cells) were transfected by calcium phosphate precipitation as described in Materials and Methods. Cotransfections included 10 µg of the indicated reporter plasmid and 200 ng of either the Sp1 expression vector pPacSp1 (+Sp1) or the pPac0 empty vector control (Sp1). Reporter plasmid constructs were the same as used in Fig. 3. Results are presented as CAT activity (cpm/µg/min) and represent the mean ± S.D. of triplicate determinations from a single experiment. Results are representative of experiments repeated three times.

Role of Sp1 Sites in Conferring Ethanol Responsiveness Depends on Promoter Context. To further assess the role of an Sp1 consensus site in ethanol responsiveness, we generated a series of CAT reporter constructs containing artificial promoters coupled to the MSV LTR. We previously showed that the MSV LTR contained in plasmid pMSV is unresponsive to ethanol, although it directs a high basal level of CAT expression in NG108-15 cells (Miles et al., 1991). The pMSV plasmid contains the 72-bp direct repeat region of the MSV LTR but does not have any consensus Sp1 sites (M.W.S. and M.F.M., unpublished results). We therefore studied whether the addition of Sp1 sites to pMSV would confer ethanol responsiveness. Constructs were made with one [pMSV(Sp1)] or three [pMSV(3Sp1)] Sp1 consensus sites from the HTLV-3 LTR inserted into a unique SstI site in pMSV. This placed the 3'-end of the Sp1 sites 5 bp from the 5'-end of the MSV TATA element. Pascal and Tjian (1991) previously showed that these Sp1 sites are capable of synergistic interaction when placed adjacent to one another. We expected, therefore, to see a greater ethanol response with construct pMSV(3Sp1).

View larger version (35K): [in this window] [in a new window]   Fig. 6.   Sp1 consensus sites confer ethanol responsiveness depending on promoter context. Plasmids containing Sp1 consensus binding sites were assessed for responses to ethanol by transient transfection analysis in NG108-15 cells. Constructs were made as described in Materials and Methods to contain the MSV LTR coupled to the proximal promoter region of MSV, hsc70, or the SV40 early promoter. In addition, Sp1 sites were contained in some constructs as indicated. Plasmid pMHsc74 m62 is identical with pMHsc74 except for a single base pair mutation in the distal Sp1 site (*Sp1). Transfections were done as described in Fig. 3. Twenty-four hours after transfection, cells were treated for 24 h with or without ethanol (100 mM) followed by lysis and assay for CAT activity. Results are expressed as absolute CAT activity (A) or as percent control (B, no ethanol treatment). Percent control results (B) are the mean ± S.E. from experiments repeated six times. Statistical significance was determined by single-group t test analysis with Bonferroni correction for multiple groups (for differences from control cells) or analysis of variance (ANOVA) with Scheffé's F test post hoc analysis to determine differences between various constructs: *P < .05 versus control (t test); P < .05 versus pMHsc74 () or pMGC6 () (ANOVA, Scheffé's post hoc).

Ethanol-Responsive Region of hsc70 Promoter Binds an Sp1-Like Protein. We performed EMSA studies to determine whether ethanol treatment of NG108-15 cells alters the abundance or DNA-binding activity of any protein factor binding to the 67/61 Sp1 consensus site. When the 74/41 region of the hsc70 promoter was used as probe, two major bands were seen on EMSA (Fig. 7A). Both of these bands were greatly reduced by competition with a 50-fold molar-excess of unlabeled probe (Fig. 7A, lanes 3 and 4), whereas oligonucleotides containing the 62 mutation (Fig. 3B) competed much less efficiently (Fig. 7A, lanes 5 and 6). Competition with unlabeled oligonucleotides containing only a consensus Sp1 site (Santa Cruz Biotechnology) was very effective in eliminating the major bands seen on EMSA (Fig. 7A, lanes 7 and 8). The most predominant EMSA band seen with NG108-15 whole-cell extracts migrated with the same mobility as a band produced with purified Sp1 (Promega) as protein source (Fig. 7A, lane 9). The more rapidly migrating band seen on EMSA with NG108-15 extracts (* in Fig. 7A) may represent either a proteolytic fragment of Sp1 or another DNA-binding protein capable of recognizing the hsc70 Sp1 consensus site. These EMSA results show that an Sp1-like binding activity present in NG108-15 cells interacts with the hsc70 promoter region required for ethanol responsiveness.

View larger version (40K): [in this window] [in a new window]   Fig. 7.   Gel mobility shift analysis of protein binding to ethanol-responsive region of hsc70 promoter. A, whole-cell lysates from control (C) or ethanol-treated (E, 100 mM, 48 h) cells were analyzed for protein binding to a 32P-probe from the 41 to 74 region of the hsc70 promoter. Results are representative of five independent experiments. Competition with 50-fold molar excess of the indicated cold DNA fragments (lanes 3-8) was used to verify specific binding. The more slowly migrating, major band is designated as Sp1 based on competition with cold consensus Sp1 oligonucleotide (lanes 7 and 8) and coincidence with a band produced by purified human Sp1 (lane 9). EMSA assays routinely contained another, more rapidly migrating band (*), which was also eliminated by competition with consensus Sp1 oligonucleotide (lanes 7 and 8). This latter band was not seen when purified Sp1 was used as a protein source (lane 9). B, EMSA assays were conducted as above except that an artificial Sp1-binding site oligonucleotide (Santa Cruz Biotechnology) was used as a probe. Competition with 50-fold cold excess of homologous Sp1 oligonucleotide or a mutated form (Santa Cruz Biotechnology) is indicated. Supershift assays with commercial antibodies (Santa Cruz Biotechnology) against Sp1 or Sp3 determined the identity of the major bands as indicated by arrows. Ethanol treatment (E) did not alter the binding activity for ethanol Sp1 or Sp3.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Our previous studies have shown that chronic ethanol exposure increases the expression of several molecular chaperones in neural cell cultures (Miles et al., 1991, 1994; Hsieh et al., 1996). Regulation of molecular chaperone expression could modify protein trafficking and contribute to synaptic reorganization that produces altered behavioral responses with chronic ethanol exposure. Furthermore, our studies in cultured neuroblastoma cells serve as a model to identify mechanisms by which ethanol regulates neuronal gene expression. Identifying how ethanol regulates molecular chaperone expression in NG108-15 neuroblastoma cells could thus have importance in understanding other ethanol-responsive gene regulation events occurring in the CNS of behaving animals. Our studies here show that a consensus-binding site for the widely used transcriptional activator Sp1 is required for ethanol induction of hsc70 transcription. We also found that other features of the proximal promoter region may restrict, in a manner not yet understood, which Sp1-site containing promoters actually respond to ethanol. Our results thus suggest a mechanism by which ethanol could cause selective changes in CNS gene expression.

Preliminary deletion studies and comparison of hsc70 promoter regions from multiple species (Fig. 1B) suggested that ethanol-responsive cis-acting sequences reside in the proximal 74 bp. Indeed, plasmid pHsc74 actually showed a larger fold-induction by ethanol than seen with pHsc2500 (Figs. 2 and 3). Importantly, pHsc74 did not respond to arsenite while maintaining a vigorous response to ethanol (Fig. 2B). This firmly dissociates the mechanism of ethanol induction from that governing response to stress protein inducers such as arsenite.

Further deletion and site-directed mutation analysis of the hsc70 promoter showed that ethanol responsiveness significantly decreased when sequences between 37 and 75 were removed or altered. In particular, point mutation of a consensus Sp1-binding site within this region greatly reduced ethanol responsiveness (pHsc74 m62). This mutation also markedly decreased induction by Sp1 in SL2 cells (Fig. 5) and decreased Sp1 binding in EMSA studies (Fig. 7A).

The ability of Sp1 sites to confer ethanol responsiveness is emphasized by the marked ethanol induction seen with pGC6 or pMGC6 (Fig. 6B), which contain the six Sp1 sites and proximal promoter/TATA box of the SV40 early promoter. However, simply adding functional Sp1 sites to a heterologous promoter (MSV) did not confer ethanol responsiveness (Fig. 6B, top). Sp1 sites inserted downstream of the MSV LTR were only responsive to ethanol when combined with the hsc70 or SV40 proximal promoter regions (Fig. 6B, middle and bottom). This suggests that the proper context of the proximal promoter region may be required before an Sp1 site-containing promoter can respond to ethanol.

However, our results regarding promoter context must be interpreted with some caution. It is possible that the pMSV(Sp1) and pMSV(3Sp1) constructs did not respond to ethanol due to improper spacing between the Sp1 sites and other promoter components. Our studies in SL2 cells showed that the Sp1 sites in these constructs were indeed functional but that ethanol responsiveness might require a particular configuration between Sp1 sites and, for example, the basal transcription apparatus. In addition, the pMSV(Sp1) and pMSV(3Sp1) constructs might have appeared unresponsive to ethanol because of high basal activities that overwhelmed any effect of ethanol. This seems less likely because some of the Hsc70 constructs (e.g., pHsc2500 and pHsc74) themselves have high basal activity. Furthermore, despite multiple experiments with several constructs with a wide range of basal activity, we never observed ethanol induction of the pMSV(Sp1) or pMSV(3Sp1) constructs. Regions of sequence homology between the hsc70 and SV40 promoters may help identify proximal promoter sequences needed for ethanol responsiveness of an Sp1-containing promoter.

The current data thus suggest that Sp1 is a crucial factor in determining ethanol regulation of the hsc70 gene. How ethanol might act to increase transcription through Sp1 is, at present, unknown. EMSA studies showed that the 67/61 Sp1 consensus site binds Sp1-like proteins present in NG108-15 cell extracts (Fig. 7A). Ethanol, however, did not alter DNA-binding activity of Sp1 or Sp3 proteins on EMSA studies (Fig. 7, A and B). Sp3 binds to the same recognition site as Sp1 and has been shown to repress Sp1-dependent transcription (Hagen et al., 1994). Ethanol must therefore increase Sp1-dependent transcription without altering Sp1 or Sp3 DNA-binding activity.

Sp1-dependent transcription has been shown to be subject to several forms of regulation. Sp1 abundance (Saffer et al., 1991; Persengiev et al., 1996), phosphorylation state (Daniel et al., 1996), and extent of O-glycosylation are all sites of regulation that can alter Sp1-dependent transcription. In particular, glycosylation is thought to increase Sp1 transcriptional activity without altering Sp1 DNA binding (Jackson and Tjian, 1988). Furthermore, Sp1 is capable of interacting with a number of other transcription factors that may increase or decrease Sp1-dependent transcription (Pugh and Tijan, 1990; Martin, 1991; Chen et al., 1994). Our data suggesting that Sp1 confers ethanol responsiveness only when paired with a specific proximal promoter region indicate that Sp1 interaction with other transcription factors or cofactors might indeed be a site of ethanol action.

Our results showing an important role for Sp1 in ethanol regulation of hsc70 gene transcription are supported by the work of other investigators. Parent et al. (1987) reported that ethanol treatment caused a 2-fold increase in expression of CAT constructs containing a 1083-bp promoter fragment from the porcine PD1 major transplantation antigen gene. This promoter region contains Sp1 consensus sites among other promoter elements. Furthermore, these investigators also showed that ethanol increased expression of plasmid pSV2CAT, which contains, among other SV40 promoter sequences, the same tandem array of Sp1 sites present in pGC6 (Fig. 6B, bottom). Furthermore, Ekblom et al. (1996) showed that ethanol increased the expression of monoamine oxidase (MAO)-B promoter constructs in SH-SY5Y neuroblastoma cells and 1242 MG glioma cells. With the 1242 MG cell line, there was a concomitant increase in Sp1-like-binding activity when using a fragment of the MAO gene in EMSAs. However, no mapping studies were done to causally link this Sp1-binding site to ethanol-induced increases in MAO-B expression.

In summary, our results indicate that a consensus Sp1-binding site is crucial for ethanol responsiveness of the hsc70 promoter. The specificity of the response may be further dictated by proximal promoter sequences or spacing configurations present in the hsc70 gene and the SV40 early promoter region. These additional factors mediating ethanol regulation of Hsc70 remain to be determined. It will be of interest to determine whether similar mechanisms regulate other ethanol-responsive genes. The identification of ethanol-responsive cis-acting promoter elements and cognate-binding proteins may provide a molecular basis for the aspects of CNS adaptation to ethanol.