|Home | About | Journals | Submit | Contact Us | Français|
The γ-aminobutyric acid (GABA) system is implicated in the neurobiological actions of ethanol, and pharmacological agents that increase the activity of this system have been proposed as potential treatments for alcohol use disorders. As ethanol has its own GABA mimetic properties, it is critical to determine the mechanism by which GABAergic drugs may reduce the response to ethanol (i.e. via an inhibition or an accentuation of the neurobiological effects of ethanol).
In the present study, we examined the ability of three different types of GABAergic compounds, the GABA reuptake inhibitor NO-711, the GABAA receptor agonist muscimol, and the GABAB receptor agonist baclofen, to attenuate the locomotor stimulant response to ethanol in FAST mice, which were selectively bred for extreme sensitivity to ethanol-induced locomotor stimulation. In order to determine whether these compounds produced a specific reduction in stimulation, their effects on ethanol-induced motor incoordination were also examined.
NO-711, muscimol, and baclofen were all found to potently attenuate the locomotor stimulant response to ethanol in FAST mice. However, both NO-711 and muscimol produced a marked increase in ethanol-induced ataxia, whereas baclofen did not accentuate this response.
These results suggest that pharmacological agents that increase extracellular concentrations of GABA and GABAA receptor activity may attenuate the stimulant effects of ethanol by accentuating its intoxicating and sedative properties. However, selective activation of the GABAB receptor appears to produce a specific attenuation of ethanol-induced stimulation, suggesting that GABAB receptor agonists may hold greater promise as potential pharmacotherapies for alcohol use disorders.
Alcoholism remains a serious health problem in the United States; however, few drugs are approved for its treatment (Miller, 2008). While abstinence is often the standard for treatment success, there is increased debate as to whether moderation should be an alternative treatment goal (Finney and Moos, 2006; Marlatt and Witkiewitz, 2002). Some approved pharmacotherapies are targeted to reduce alcohol (ethanol) consumption by a direct pairing with ethanol, whereas others target craving during abstinence. For example, disulfiram reduces drinking by inducing adverse physiological symptoms when combined with ethanol (Suh et al., 2006). Naltrexone is prescribed partly with the intent of increasing number of days abstinent, but when combined with ethanol, it has been shown to reduce the number of drinks consumed, possibly by decreasing the positive, rewarding effects of ethanol (O’Malley et al., 1996; Sinclair, 2001; Volpicelli et al., 1992). However, Swift and colleagues (1994) reported an enhancement of subjective intoxication ratings, including sedation, when naltrexone was combined with ethanol. This provides an alternative explanation for reduced consumption, namely increased intoxication and sedation effects, effects which might be seen after consumption of higher ethanol doses.
Ethanol, among its many effects, has modulatory actions on GABA systems, including alterations in GABA cell firing and synaptic release (Gallegos et al., 1999; Roberto et al., 2003; Siggins et al., 2005; Xiao et al., 2007; Zhu and Lovinger, 2006) and increases in GABAA receptor function (Allan and Harris, 1987; Suzdak et al., 1986; Wallner et al., 2003). Many of the behavioral effects of ethanol, including sensitivity to its locomotor stimulant, sedative, incoordinating, and reinforcing effects can be attributed, in part, to alterations in GABA systems (Boehm et al., 2006; Chester and Cunningham, 2002; Dar, 1996; Hoffman et al., 1987; Koob, 2004; Martz et al., 1983; Phillips and Shen, 1996; Shen et al., 1998). Benzodiazepines, which share with ethanol the ability to positivity modulate GABAA receptors, are efficacious in the treatment of ethanol withdrawal. However, recent attention has focused on the use of antiepileptics, including topiramate, γ-vinyl GABA, and tiagabine for the treatment of alcoholism. These drugs, among other actions, enhance GABAA receptor function or increase extracellular GABA concentrations (Davies, 1995; Fink-Jensen et al., 1992; White et al., 2000). In addition, the use of drugs that target the GABAB receptor, such as baclofen, is being explored. Both GABAA and GABAB receptor related drugs have been shown to reduce ethanol consumption in preclinical and clinical studies (Addolorato et al., 2000; Anstrom et al., 2003; Colombo et al., 2000; Johnson et al., 2003; Nguyen et al., 2005; Wegelius et al., 1993). However, reduced consumption could be due to the accentuation of a competing behavior, such as increased intoxication or motor incoordination, rather than an attenuation of the reinforcement experienced when ethanol is consumed. Such accentuated motor side effects would pose significant patient safety concerns.
To explore the possibility that compounds with GABA mimetic effects attenuate alcohol effects by inducing competing motor behaviors, we examined the effects of GABAergic compounds on ethanol-induced locomotor activation and motor coordination. We hypothesized that the GABA mimetic drugs would attenuate ethanol-induced stimulation, but that they would also enhance ataxia. This would suggest that the reduced stimulation is due to the additive effect of the GABAergic drugs with ethanol, leading to the expression of competing behaviors, namely ataxia. Sensitivity to the behaviorally stimulating and incoordinating effects of ethanol are of additional interest because increased sensitivity to ethanol-induced stimulation and decreased sensitivity to motor incoordination are correlates of ethanol consumption history (Holdstock et al., 2000; King et al., 2002) and familial risk for alcoholism (Newlin and Thomson, 1991, 1999; Schuckit and Smith, 2000, 2001).
We chose the FAST and SLOW mouse lines for these studies. These mice were created by selectively breeding for extreme sensitivity and insensitivity, respectively, to the locomotor stimulant effects of ethanol (Crabbe et al., 1987; Phillips et al., 1991). They therefore represent a genetic model of differential sensitivity to the stimulant effects of ethanol. However, they also differ dramatically in sensitivity to the motor incoordinating and sedative effects of ethanol (with FAST mice less sensitive to these effects; Phillips et al., 2002; Shen et al., 1996), a difference that grew larger over the course of selection (Phillips et al., 2002), providing strong evidence for common genetic regulation of these ethanol effects. Further, FAST mice have been found to consume larger amounts of ethanol than SLOW mice (Risinger et al., 1994). Thus, the FAST lines, which exist as two replicates (FAST-1 and FAST-2) appear to model the high stimulant sensitivity, low sedative sensitivity and higher drinking reminiscent of at least some individuals who are prone to alcoholism.
We first examined the effect of the GABA transporter inhibitor NO-711, which increases extracellular GABA concentrations (Fink-Jensen et al., 1992), on ethanol-induced locomotion in FAST and SLOW mice. We predicted that GABA receptor activation induced by NO-711 would attenuate ethanol stimulation in FAST mice by acting additively with ethanol to accentuate GABAergic activity. We also predicted that NO-711 would accentuate the depressant response to ethanol normally seen in SLOW mice. The relative contributions of GABAA versus GABAB receptors to attenuation of stimulation were then considered by using drugs that specifically interact with these receptor subtypes. Only FAST mice were used in these studies, due to the desire to see possible effects on both stimulation and motor incoordination (SLOW mice do not exhibit ethanol-induced stimulation at any dose; Palmer et al., 2002a). We predicted that the GABAA receptor agonist muscimol would attenuate the stimulant response to ethanol in FAST mice, but would also enhance the motor incoordinating effects of ethanol. We also examined the effect of baclofen, a GABAB receptor agonist, which has been previously shown to attenuate the stimulant response to ethanol in FAST mice (Boehm et al., 2002; Shen et al., 1998). Because the GABAB receptor agonist baclofen has muscle relaxant properties like ethanol (Nevins et al., 1993), we postulated that baclofen would also accentuate the incoordinating effects of ethanol.
The FAST and SLOW selected lines were selectively bred, in replicate, for extreme sensitivity (FAST-1, FAST-2) and insensitivity (SLOW-1, SLOW-2) to the locomotor stimulant effects of 2 g/kg ethanol. Details specific to the selection process have been published (Crabbe et al., 1987, 1988; Phillips et al., 1991, 2002; Shen et al., 1995). The most recent test of the ethanol dose-response function in these mice (~20 generations since selection was relaxed) showed that FAST mice were significantly different from SLOW mice in their locomotor response to multiple ethanol doses (i.e., 1.5, 2, 2.5, and 3 g/kg). Mice used in these experiments were reared with the dam and sire until 20–22 days of age, when they were isosexually housed in polycarbonate [28 × 18 × 13 cm (l × w × h)] cages, 2–4 per cage, with littermates or with non-littermates (to avoid single housing) of the same genotype and age range. Subjects were maintained on a 12:12-h light-dark cycle (lights on at 0600) at 21 ± 2° C with food (Purina Laboratory Rodent Chow #5001; Purina Mills, St. Louis, MO) and water available ad libitum except during behavioral testing. The age, number and generations of mice used for each study are presented in the figure legends.
Eight locomotor activity detection monitors measuring 40 × 40 × 30 cm (l × w × h) (AccuScan Instruments, Inc., Columbus, OH) were each housed in light-proof, sound-attenuating cabinets (Flair Plastics, Portland, OR). Each cabinet was illuminated by an 8-W fluorescent white light and a fan was mounted on the inside back wall, providing ventilation and background noise. Movement within the monitor was recorded by two sets of eight infrared beams mounted 2 cm above the test chamber floor at right angles to one another, with detectors mounted on the opposite sides. Beam interruptions were automatically recorded and translated to horizontal distance traveled (in cm) by AccuScan software.
Motor incoordination was measured on two AccuRotor Rota Rods (AccuScan Instruments, Columbus, OH). The rotarod consisted of a 6.3 cm diameter dowel covered with 320 grit wet/dry sandpaper that was divided into four lanes by white plexiglass round discs. The dowel was located 63 cm above a thick layer of sawdust bedding, with a photobeam detector that recorded latency to fall from the dowel in sec.
All drugs were dissolved in 0.9% physiological saline (Baxter Healthcare Corporation, Deerfield, IL) on the day of testing. (±)-Baclofen, muscimol, and NO-711 HCl were obtained from Sigma (St. Louis, MO) and injected i.p. at a volume of 10 ml/kg. Ethanol was obtained from Pharmco Products (Brookfield, CT) and was diluted from a 100% stock to a 20% v/v solution in saline and injected i.p.
All testing occurred between 0800 and 1600 h. At the end of each experiment, all subjects were euthanized by carbon dioxide asphyxiation. These procedures were approved by the Portland Veterans Affairs Medical Center Institutional Animal Care and Use Committee and complied with the National Institutes of Health Guidelines for the Care and Use of Laboratory Animals (1986).
Mice were moved in their home cages to the procedure room and left undisturbed to acclimate for 45–60 min prior to testing. Subjects were not habituated to the activity monitor or to injections prior to locomotor testing, consistent with the selection procedure (Crabbe et al., 1988; Phillips et al., 1991), as well as previous experiments using GABA mimetic compounds in FAST mice (Palmer et al., 2002a,b; Phillips et al., 1992; Shen et al., 1998). Mice were weighed, injected with the pretreatment drug (saline, NO-711, muscimol, or baclofen), and placed individually into holding cages. At the conclusion of the pretreatment interval (specified below), each subject was injected with saline or ethanol and placed into the activity monitor. Activity was recorded for 30 min in 5-min time bins. At the conclusion of testing, blood samples (20 µl) were collected from the peri-orbital sinus of ethanol-treated mice for the determination of blood ethanol concentration (BEC), and processed as described by Boehm et al. (2000). BECs were determined using gas chromatography (Agilent 6890) with flame ionization detection.
This procedure was adapted from Rustay and colleagues (2003a,b). Mice were moved to the procedure room, weighed, and left undisturbed for approximately 30–45 min. Each mouse was then placed onto the dowel, which rotated at a fixed speed of 3 RPM, and up to five criterion trials were conducted. Subjects met criterion if they successfully remained on the rod for 180 sec. Only one successful trial was required to meet criterion. Afterwards, subjects were returned to their home cages to await testing (for a minimum of 30–45 min), at which time they were pretreated (saline, NO-711, muscimol, or baclofen) and placed in individual holding cages. After the pretreatment interval, subjects were injected with saline or ethanol and placed immediately on the rotating dowel. Three trials were conducted at the T0 time point (immediately after ethanol administration) and at the T10 time point (10 min after ethanol administration), with a 180 sec maximum duration upon the rotarod per trial. Latency to fall from the dowel in sec was automatically recorded.
FAST and SLOW (replicate 1 and 2) male mice were injected with saline or NO-711 (1.25, 2.5, 5, or 7.5 mg/kg) 30 min prior to injection with saline or 2 g/kg ethanol. The NO-711 doses and pretreatment interval were adapted from Dalvi and Rodgers (1996), while the 2 g/kg ethanol dose was the dose used for the majority of selective breeding generations of the FAST and SLOW lines (Crabbe et al., 1988; Phillips et al., 1991). A follow-up experiment used the same procedures and doses of NO-711 to assess the effects of NO-711 on the locomotor response to a lower dose of ethanol (1 g/kg) which was previously found to elicit only a modest stimulant response in FAST mice (Palmer et al., 2002a).
The effect of NO-711 on the incoordinating effects of ethanol was examined in male FAST-2 mice administered saline or NO-711 (5 and 7.5 mg/kg) 30 min prior to an injection of saline or ethanol (1.2 g/kg). These doses of NO-711 were the most effective at reducing ethanol-induced stimulation in FAST-2 mice in Experiment 1. The dose of ethanol chosen for this experiment was one known to have only minor ataxic effects, thereby allowing for increased ataxia to be observed when ethanol was given in combination with NO-711 (JC Crabbe, personal communication; Rustay et al., 2003a,b). Only FAST-2 mice were used for this study due to their greater availability, and because comparable results were obtained in the replicate lines in Experiment 1.
To determine whether GABAA receptor activation alone would have effects on ethanol-induced locomotor stimulation similar to the effects of NO-711, male FAST-2 mice were injected with saline or muscimol (0.5, 1, 1.5, or 2 mg/kg) 10 min prior to an injection of saline or 2 g/kg ethanol (doses and pretreatment interval adapted from Shen et al., 1998). Due to their greater stimulant response to ethanol (Palmer et al., 2002a) and their greater availability, only FAST-2 mice were used for this study.
To determine whether muscimol also accentuated ethanol-induced ataxia, male FAST-2 mice were administered saline or muscimol (1 and 1.5 mg/kg) 10 min prior to saline or 1.2 g/kg ethanol. The muscimol doses used were the most effective at reducing ethanol-induced stimulation in Experiment 3 in the absence of gross motor side-effects.
In a previous study with FAST mice, the GABAB receptor agonist baclofen was found to attenuate the locomotor stimulant response to 2 g/kg ethanol (Shen et al., 1998). The purpose of this experiment was to examine the possibility that baclofen reduced stimulation by accentuating ethanol-induced motor incoordination. In the initial study, male FAST-2 mice were tested following saline or baclofen (1.25, 2.5, or 5 mg/kg) administered 15 min prior to saline or 1.2 g/kg ethanol. Due to unpredicted results, the study was replicated in male FAST-1 mice.
To enhance interpretation of the effects of baclofen on ethanol-induced rotarod ataxia, the effect of baclofen on stimulation to a low, sub-stimulating dose of ethanol was examined in FAST mice. FAST-1 and -2 mice were administered saline or baclofen (0.625, 1.25, 2.5, or 5 mg/kg) 15 min prior to an injection of saline or ethanol (1 g/kg).
For the locomotor activity studies, repeated measures analyses of variance (ANOVAs) were performed to determine whether there was an interaction with time, prompting separate analyses to determine the time period in which the drug effects were most robust. We determined from these initial analyses that the first 10-min test period best described the drug effects in these studies. Further, this is the time when stimulation to ethanol is most profound (Shen et al., 1995). Data for this time period were then analyzed by factorial ANOVA. Across the multiple studies, several different categorical variables were included in the analysis, including line, replicate, pretreatment dose, and ethanol dose. Interactions of three or more factors were followed up by two-factor ANOVAs as appropriate to further parse the data, and by simple main effects analyses and Newman-Keuls mean comparisons to examine significant two-way interactions and main effects, respectively. For the rotarod ataxia experiments, nonparametric analyses were conducted because the data were not normally distributed (due to a ceiling criterion). Individual Kruskal-Wallis tests were performed to examine the effect of drug pretreatment on saline- and ethanol-responding. Significant effects of pretreatment, as well as analyses of ethanol effects at each pretreatment dose, were examined by Bonferroni-corrected Mann-Whitney U comparisons. Significance levels were set at α≤0.05. All analyses were conducted with the Statistica 6.1 software package (StatSoft, Inc., Tulsa, OK).
In a separate analysis of data from FAST and SLOW mice that had been pretreated with NO-711 prior to saline administration, FAST and SLOW mice were found to differ in their locomotor response to NO-711 alone (Fig 1). This genetic correlation indicates that some common genes, and therefore mechanism(s), influence sensitivity to ethanol and to this GABA transporter inhibitor. NO-711 produced a significant decrease in the locomotor activity of SLOW mice while not significantly altering the locomotor activity of FAST mice, compared to their activity after saline injection (Fig 1).
Results of a repeated measures ANOVA supported these characterizations. Because replicate interacted with time and NO-711 dose [F4, 190=2.6, p<0.05], separate analyses were performed on data from each replicate. For replicate 1 data, there was a significant line X NO-711 dose interaction [F4, 92=4.8, p<0.01], and post hoc analyses identified significant differences in locomotor response to 2.5, 5, and 7.5 mg/kg NO-711 between FAST-1 and SLOW-1 mice. Moreover, NO-711 (2.5, 5, and 7.5 mg/kg) significantly reduced the activity of SLOW-1 mice, while having no effect on the activity of FAST-1 mice. For replicate 2 data, there was also a significant line X NO-711 dose interaction [F4, 98=3.2, p<0.05]. While interpretation is complicated by a line difference between FAST-2 and SLOW-2 mice in spontaneous activity, the lines were found to differ in response to 5 and 7.5 mg/kg NO-711. Moreover, the activity of SLOW-2 mice was significantly reduced by the 7.5 mg/kg dose; however, NO-711 did not significantly reduce activity in FAST-2 mice.
In FAST mice, pretreatment with NO-711 dose-dependently abolished the locomotor stimulant response to the selection-dose of ethanol (2 g/kg), while attenuating stimulation to 1 g/kg ethanol in FAST-1, but not FAST-2, mice. Conversely, NO-711 accentuated the locomotor depressant response to 2 g/kg ethanol in SLOW mice (Fig 2). The following statistical analyses supported the above-stated characterizations.
Data for the 1 and 2 g/kg ethanol doses were collected in separate studies. Our initial study examined the effect of NO-711 on activity after saline or 2 g/kg ethanol in both FAST and SLOW mice. To further test the hypothesis that ethanol and NO-711 have additive effects, a separate study was then performed examining the effect of NO-711 on activity after saline or 1 g/kg ethanol, a sub-maximally stimulating dose in FAST mice only; there is no dose of ethanol that induces stimulation in SLOW mice. For the purpose of probing the effect of NO-711 across the two ethanol doses, data for FAST mice from these two studies were combined in a single analysis. Activity levels of animals treated with NO-711 prior to saline were not found to significantly differ between the two studies, thus data from the two control groups were combined for this analysis (open symbols in Fig 2, panels a and b).
Due to an interaction of replicate with NO-711 dose and ethanol dose [F8, 417=2.3, p<0.05], data from the two replicate FAST lines were considered in separate analyses. For FAST-1 mice (Fig 2a), there was a significant NO-711 dose X ethanol dose interaction [F8, 203=3.8, p<0.001]. The 1 and 2 g/kg ethanol doses unexpectedly induced significant stimulant responses of similar magnitude in this line of mice. While NO-711 pretreatment reduced the stimulant response to both ethanol doses, only the highest dose of NO-711 attenuated the stimulant response to 1 g/kg ethanol, while moderate to high doses of NO-711 (2.5–7.5 mg/kg) attenuated stimulation to 2 g/kg ethanol. There was no effect of NO-711 on saline activity.
For FAST-2 mice (Fig 2b), there was also a NO-711 dose X ethanol dose interaction [F8, 214=7.3, p<0.001]. The 1 and 2 g/kg ethanol doses significantly increased activity; however, the stimulant response was dose-dependent in this mouse line. NO-711 pretreatment (5 and 7.5 mg/kg) attenuated the stimulant response to 2 g/kg, whereas NO-711 pretreatment did not alter the more modest stimulant response to 1 g/kg ethanol. There was no effect of NO-711 on locomotor activity after saline treatment.
SLOW mice were examined only at the 2 g/kg ethanol dose. Replicate interacted with NO-711 dose and ethanol dose [F4, 182=3.7, p<0.01]. For SLOW-1 mice (Fig 2c), there was a NO-711 dose X ethanol dose interaction [F4, 92=6.4, p<0.001]. Consistent with their selection response, 2 g/kg ethanol induced a locomotor depressant response in these mice. There was a significant effect of NO-711 on activity after both saline and ethanol treatment, with dose-dependent effects in both cases (see Fig 2). For SLOW-2 mice (Fig 2d), there were main effects of NO-711 dose [F4, 90=15.6, p<0.001] and ethanol dose [F1, 90=45.8, p<0.001], but no significant interaction of these two factors, indicating that NO-711 had similar effects on the activity of saline- and ethanol-treated mice. Thus, in general, the effect of both ethanol and NO-711 in SLOW mice was to reduce locomotor behavior.
Analysis of BEC data obtained from samples taken after the 30-min activity test revealed no effect of NO-711, and no line or replicate differences. The mean (± SEM) BEC for the genotypes and NO-711 dose groups combined was 0.68 ± 0.02 mg/ml after 1 g/kg ethanol and 1.32 ± 0.03 mg/ml after 2 g/kg ethanol at the 30-min time point.
To examine whether the attenuation of ethanol-induced locomotor stimulation in FAST mice by NO-711 could be due to an accentuation of ethanol-induced motor incoordination, rotarod ataxia data were collected. As shown in Fig 3, NO-711 significantly accentuated the ataxic effects of ethanol. Data from four mice (5.8 % of all mice tested) were excluded from the study because the mice did not meet the 180 sec baseline rotarod performance criterion. Analyses of latency to fall from the rod across the three trials at T0 (immediately after ethanol administration) and T10 (10 min post-ethanol administration) revealed no significant three-way interaction of trial, NO-711 dose, and ethanol dose. Average latency to fall at each time point adequately reflected the results and was therefore used as the dependent variable.
For the T0 time point (Fig 3a), Kruskal-Wallis analyses found no significant effect of NO-711 pretreatment on latency to fall after saline; however, NO-711 did significantly decrease latency to fall after ethanol [H2, N=33=12.8, p<0.01], with increased ataxia at 5 and 7.5 mg/kg NO-711 (combined with ethanol) as compared to ethanol alone. In Bonferroni-corrected pairwise comparisons, ethanol was found to significantly decrease latency to fall from the rod at all NO-711 doses (including 0 mg/kg).
By the T10 time point, motor incoordinating effects of ethanol and NO-711 alone had abated, yet there was an even larger enhancement of motor incoordination when these two drugs were combined (Fig 3b). Again, there was a significant effect of NO-711 on latency to fall after ethanol [H2, N=33=12.7, p<0.01], with a significant decrease in latency scores after treatment with both 5 and 7.5 mg/kg NO-711. NO-711 pretreatment did not affect latency to fall after saline. Furthermore, Bonferroni-corrected pairwise comparisons revealed significant differences between ethanol- and saline-treated mice after treatment with 5 and 7.5 mg/kg NO-711, but not 0 mg/kg NO-711. Overall, these data indicate increased ataxia with the combination of NO-711 and ethanol.
As shown in Fig 4, pretreatment with the GABAA receptor agonist muscimol significantly attenuated the locomotor stimulant response to ethanol in FAST-2 mice, an effect similar to that observed for NO-711 in Experiment 1. A significant muscimol dose X ethanol dose interaction was found [F4,109=25.2, p<0.001]. Muscimol altered the locomotor response to both saline and ethanol, with a modest locomotor stimulant response induced by 1 and 1.5 mg/kg muscimol in saline-treated mice. However, in ethanol-treated mice, the stimulant response to ethanol was significantly reduced by the three highest doses of muscimol. There was no effect of muscimol pretreatment on BEC determined from samples collected after the 30 min activity test; mean BEC (± SEM) for the combined dose groups was 1.61 ± 0.05 mg/ml.
Similar to NO-711, muscimol dramatically enhanced the motor-incoordinating effects of a low dose of ethanol in FAST-2 mice (Fig 5). Data from four mice (5.3 %) were excluded from the study because the mice were unable to meet the baseline performance criterion. Analyses at both the T0 and T10 time points revealed no interaction of trial, muscimol dose, and ethanol dose, and therefore an average of the 3 trials at each time point was used as the dependent variable.
Immediately after ethanol administration (T0), muscimol significantly decreased latency to fall from the rod after ethanol [H2, N=36=17.5, p<0.001], with an increase in ataxia at both 1 and 1.5 mg/kg muscimol (Fig 5a). While muscimol did affect latency to fall after saline [H2, N=35=8.2, p<0.05], a significant decrease in latency was only observed at the highest muscimol dose (1.5 mg/kg). With a muscimol-induced effect on motor coordination, there was only a significant difference in latency to fall between ethanol- and saline-treated mice at 1 mg/kg muscimol; ethanol alone did not significantly alter latency to fall.
As seen in Fig 5b, a similar pattern of results was seen 10 min post-ethanol treatment (T10). Kruskal-Wallis analyses revealed significant effects of muscimol on latency to fall after saline [H2, N=35=15.1, p<0.001] and ethanol [H2, N=36=15.3, p<0.001]. While both 1 and 1.5 mg/kg muscimol, in combination with ethanol, increased motor incoordination as compared to ethanol alone, it was only the 1.5 mg/kg dose of muscimol that had ataxic effects on its own. Bonferroni-corrected pairwise comparisons highlighted a significant difference in latency scores between saline- and ethanol-treated mice at 1 mg/kg muscimol, but not at 0 or 1.5 mg/kg muscimol.
As shown in Fig 6a, pretreatment with baclofen did not accentuate ethanol-induced motor incoordination in FAST-2 mice, an effect which is in stark contrast to the effects of NO-711 and muscimol. Data from six mice (7.2 %) were excluded from the study because the mice were unable to meet the baseline performance criterion. Analyses did not reveal any interaction of baclofen dose and ethanol dose with trial, so the data were analyzed as a T0 average and a T10 average.
At the T0 time point, Kruskal-Wallis tests revealed no effect of baclofen pretreatment on latency to fall from the rod after saline or ethanol. Additionally, saline- and ethanol-treated FAST-2 mice only differed in latency to fall at the 1.25 mg/kg dose of baclofen. At the T10 time point, there was no significant effect of baclofen on ethanol or saline responding.
We had predicted that baclofen would accentuate ethanol-induced ataxia; it did not. To confirm these results, the study was repeated using FAST-1 mice, and as shown in Fig 6b, the results were replicated–baclofen did not accentuate ethanol-induced motor incoordination in this line either. Data from one mouse (1.3 %) were excluded from the study because the mouse did not meet the performance criterion. Immediately after ethanol administration (T0), there was again no effect of baclofen pretreatment on latency to fall after saline or ethanol in FAST-1 mice. At the T10 time point, while there was no effect of baclofen on latency to fall after saline treatment, there was a significant effect of baclofen on latency to fall after ethanol [H3, N=38=9.2, p<0.05]. Bonferroni-corrected post-hoc comparisons revealed that there was actually a significant increase in latency to fall after ethanol with 2.5 mg/kg baclofen as compared to 0 mg/kg baclofen. These results confirmed those obtained using FAST-2 mice that baclofen, in this procedure, does not accentuate ethanol-induced motor incoordination.
The striking contrast in motor coordination results seen for NO-711 and muscimol versus baclofen was not predicted. While our laboratory has previously found that peripherally administered baclofen attenuates the locomotor stimulant response to ethanol in FAST mice (Shen et al., 1998), the ethanol dose used was 2 g/kg, a dose higher than that used for the rotarod study (1.2 g/kg). Experiment 6 was designed to determine whether baclofen would reduce stimulation to a lower dose of ethanol, one more comparable to that used to measure combined drug effects on ataxia. Since baclofen did not accentuate ataxia produced by a relatively low dose of ethanol, a reduction by baclofen of the stimulation produced by this low dose of ethanol would not likely be due to altered coordination. As shown in Fig 7, baclofen dose-dependently reduced the locomotor stimulant response to 1 g/kg ethanol in FAST-1 mice, but did not reduce the stimulant response to this dose of ethanol in FAST-2 mice. A replicate X baclofen dose X ethanol dose interaction [F4, 254=3.2, p<0.05] prompted a separate analysis of the data for each replicate line. For FAST-1 mice, there was a significant interaction of baclofen dose and ethanol dose [F4, 139=2.6, p<0.05]. A stimulant response to 1 g/kg ethanol was observed in these mice, which was significantly reduced by the highest dose of baclofen (5 mg/kg). In contrast, ethanol had stimulant effects in FAST-2 mice [F1, 115=81.5, p<0.001], but there was no effect of baclofen or significant interaction of baclofen dose and ethanol dose. BECs were comparable across baclofen pretreatment and replicate line groups in samples collected after the 30 min activity test. Mean BEC (± SEM) for the combined replicate line and dose groups was 0.55 ± 0.04 mg/ml.
FAST and SLOW mice differed in sensitivity to the effects of the GABA transporter inhibitor NO-711 on locomotor activity, adding evidence for common genetic regulation of sensitivity to GABA mimetic drugs and ethanol. While activation of GABA systems, either via an indirect agonist (NO-711) or via receptor specific agonists (muscimol or baclofen), virtually eliminated the locomotor stimulant response to ethanol, there appears to be a receptor subtype-specific behavioral mechanism for this effect. The attenuation of ethanol-induced locomotor stimulation by both NO-711 and muscimol was accompanied by a dramatic increase in the motor incoordinating effects of even a low dose of ethanol. In contrast, the GABAB receptor agonist baclofen attenuated stimulation in the complete absence of an increase in ethanol-induced motor incoordination. These results emphasize that the manner in which GABAergic drugs attenuate ethanol effects should be carefully considered. The most parsimonious explanation for the results described herein is that GABA mimetic drugs that enhance GABAA receptor function act additively with ethanol to shift the behavioral response toward greater intoxication. This may be manifest behaviorally through a reduction in stimulation by the enhancement of a competing behavior, i.e., motor incoordination. Activation of GABAB receptors, however, may produce a more selective attenuation of the stimulant effects of ethanol, reducing stimulation without severe behavioral side effects. The additive effects of NO-711 or muscimol with ethanol could not be predicted from their effects on saline-treated mice alone; doses of all drugs tested were chosen to have minimal effects in saline-treated mice. Further, comparable mean responses between saline-treated and ethanol-treated mice that have been pretreated with a putative pharmacotherapeutic drug cannot be assumed to indicate the absence of a behavioral effect of the drug combination.
The FAST and SLOW mouse lines serve as a model to study the genetic and neurobiological mechanisms underlying acute ethanol sensitivity. Previous studies have implicated a role for both GABAA and GABAB receptors in the selection phenotype (Palmer et al., 2002a,b; Phillips et al., 1992; Shen et al., 1998). NO-711 (also known as NNC-711) is a highly selective and specific noncompetitive antagonist of the GAT1 GABA transporter subtype, which is generally characterized as a neuronal transporter (Minelli et al., 1995; Suzdak et al., 1992). Inhibition of GAT1 leads to increases in extracellular GABA concentrations (Fink-Jensen et al., 1992); therefore NO-711 acts as an indirect agonist of both GABAA and GABAB receptors and likely enhances the tonic inhibition of target neurons. In the current study, SLOW mice were more sensitive to the locomotor depressant effects of NO-711, which is in accord with their enhanced sedative sensitivity to other GABAA and GABAB receptor agonists and positive modulators (Palmer et al., 2002a,b; Shen et al., 1998). Ethanol is hypothesized to have GABA mimetic effects (Allan and Harris, 1987; Crisswell and Breese, 2005; Roberto et al., 2003; Wallner et al., 2003), and the line differences to a wide array of GABAergic drugs suggest that FAST and SLOW mice differ with regard to some aspect of GABA system function. Our laboratory has preliminary evidence to support this hypothesis, showing that GABAA receptor inhibitory postsynaptic currents (IPSCs) and GABAB receptor function in ventral midbrain tissue are greater in SLOW than in FAST mice (unpublished results). However, further studies are required to determine the cellular mechanisms mediating this line difference, and to address the GABAA receptor subunit-specific nature of ethanol effects that has become somewhat controversial (Borghese and Harris, 2007; Borghese et al., 2006; Hanchar et al., 2005; Korpi et al., 2007; Olsen et al., 2007; Wallner et al., 2003).
Ethanol has a similar behavioral profile to that of several GABAA receptor agonists and positive modulators, including the ability to induce behavioral stimulation (e.g., see Fig 4, showing stimulation to muscimol in FAST mice). Therefore, several groups have hypothesized that ethanol-induced increases in GABAA receptor function may contribute to ethanol-induced locomotor stimulation (Liljequist and Engel, 1982; Palmer et al., 2002a,b; Phillips et al., 1992; Phillips and Shen, 1996). Given these findings, it is peculiar that activation of GABAA receptors attenuates ethanol stimulation in FAST mice and other mouse strains (Biswas and Carlsson, 1978; Broadbent and Harless, 1999; Liljequist and Engel, 1982), whereas GABAA receptor antagonism does not (Liljequist and Engel, 1982; Shen et al., 1998). The attenuation of stimulation by agonists could be due to a specific blockade of a mechanism involved in ethanol’s stimulatory effects in the central nervous system. Alternatively, it could be due to an additive effect of two GABAergic agonists (i.e., ethanol and the other GABAergic drug), resulting in a shift of the behavioral response towards motor incoordination or intoxication, an effect likely to occur with increased GABA levels. Our data support an additive effect of ethanol and the GABA agonists. Although our data do not directly address mechanism, the genetic correlations in FAST and SLOW mice for responses to ethanol and GABAergic drugs are strongly supportive of a common mechanism involving GABA.
In FAST mice, NO-711 attenuated the locomotor stimulant response to ethanol without significantly altering the activity of saline-treated mice. This lack of an effect of NO-711 on spontaneous activity could suggest that the attenuation of ethanol stimulation was not due to a general reduction in motor activity. Analysis of the activity data in 1-min intervals (data not shown) showed an early reduction in ethanol-stimulated activity in NO-711 pretreated mice that remained reduced throughout the 10 min activity period analyzed. This could have been due to a direct inhibition of the stimulatory effects of ethanol, or to an enhancement of the motor incoordinating effects of ethanol. Because the locomotor activity procedure is not a sensitive measure of ataxia (i.e., mice can locomote even when moderately ataxic), a rotarod ataxia test was performed. Our rotarod ataxia data showed a large increase in ethanol-induced ataxia with NO-711, as well as ataxic effects of NO-711 alone. Therefore, a reduction in ethanol-induced locomotor stimulation in NO-711 pretreated mice likely occurred because of the accentuation of a competing behavior, namely ataxia. When NO-711 was combined with a lower dose of ethanol (1 g/kg), we observed a reduced ability of NO-711 to attenuate stimulation in FAST-1 mice, and a lack of attenuation of stimulation in FAST-2 mice. This may have occurred because the combination of NO-711 with a lower dose of ethanol (1 g/kg) did not reach a necessary threshold for ataxia, thereby limiting the ability of NO-711 to attenuate stimulation. In SLOW mice, the combination of NO-711 and ethanol further enhanced the locomotor depressant response normally seen to ethanol, as would be predicted if the compounds in combination act to shift the behavior towards greater intoxication.
The results of the experiment with muscimol, but not baclofen, almost entirely mimicked the results for NO-711, suggesting that the predominant mechanism for NO-711’s effects is activation of GABAA receptors. As NO-711 increases extracellular concentrations of GABA (Fink-Jensen et al., 1992), it should be increasing both GABAA and GABAB receptor function. Differences in brain distribution for these two receptor subtypes (which are unknown for these mouse lines) or differences in the behavioral effects that occur with the activation of GABAA vs. GABAB receptors may explain what appears to be a predominantly GABAA receptor-mediated effect of NO-711 on ethanol-induced ataxia. A GABAA receptor-mediated mechanism by which NO-711 and muscimol, in combination with ethanol, increases ataxia (and thereby decreases stimulation) is consistent with previous work demonstrating that GABAA receptor agonists accentuate the motor-incoordinating effects of ethanol (Dar, 2006; Martz et al., 1983), while GABAA receptor blockade results in a reduction of ethanol-induced ataxia (Dar, 2006; Hoffman et al., 1987; Suzdak et al., 1988).
The current data are complementary to recent data in humans showing that the GAT1 inhibitor tiagabine, in combination with ethanol, did not attenuate ethanol-induced activation of limbic regions, but instead produced a large depression of cerebellar activity (possibly related to the increase in ataxia seen in the current study) (Fehr et al., 2007). These data also mimic the effect of NMDA receptor antagonists on ethanol-induced motor behaviors. The NMDA receptor antagonist MK-801 eliminated the stimulant response to ethanol in mice and, at higher doses, attenuated ethanol-induced locomotor sensitization. This effect was accompanied by an increase in ethanol-induced motor incoordination (Meyer and Phillips, 2003a). As ethanol inhibits NMDA receptors (Dildy-Mayfield and Leslie, 1991), these data suggest that MK-801 and ethanol act additively to induce a shift in the behavioral response to ethanol towards greater intoxication. Similar to the current findings, this suggests that pharmacotherapies that act in a manner similar to ethanol may induce their desired behavioral effect via a potentiation of an undesired ethanol effect. However, given in the absence of ethanol, they could be considered replacement therapies, reducing ethanol consumption by partial substitution for ethanol.
That the GABAB receptor agonist baclofen had no effect on ethanol-induced motor incoordination was surprising because baclofen has potent anti-spastic and muscle relaxant properties, a trait it shares with ethanol (Bowery, 2006; Nevins et al., 1993). Our laboratory has previously found that both peripheral and central baclofen administration attenuated ethanol-induced locomotor stimulation in FAST mice (Boehm et al., 2002; Shen et al., 1998), an effect also observed in other mouse strains (Broadbent and Harless, 1999; Chester and Cunningham, 1999; Humeniuk et al., 1993). As central administration would be less likely to induce muscle relaxation, this may suggest a role for GABAB receptors in ethanol stimulation that is independent of the accentuation of ataxia. This selective reduction in stimulation could plausibly be mediated by several different neuronal mechanisms. First, activation of GABAB receptors has been found to decrease dopamine cell firing in the ventral midbrain (Erhardt et al., 2002; Mueller and Brodie, 1989), and the dopamine response to several drugs of abuse, including amphetamine, cocaine, morphine, and nicotine (Brebner et al., 2005; Fadda et al., 2003). It is likely that baclofen attenuates ethanol-induced locomotor stimulation through a similar mechanism (a hypothesis that we are currently testing using microdialysis). This would likely result in a reduction in stimulation in the absence of heightened ataxia. Another possibility is that baclofen may inhibit the effect of ethanol at GABAA receptors. In recordings from rat hippocampus, baclofen blocked the ethanol-induced potentiation of GABAA receptor function, likely due to a presynaptic effect of GABAB receptors (Ariwodola and Weiner, 2004). If this occurs in other brain regions, including the mesolimbic dopamine pathway, it is possible that baclofen would attenuate ethanol-induced locomotor stimulation by blocking the effect of ethanol at the GABAA receptor, thereby reducing the stimulatory effects of ethanol without accentuating ethanol-induced motor incoordination.
Baclofen was able to attenuate the stimulant effect of a very low dose of ethanol (1 g/kg) in FAST-1 but not FAST-2 mice. This result is in contrast to the ability of baclofen to attenuate the stimulant response to a higher, and more stimulatory, dose of ethanol (2 g/kg) in both FAST-1 and FAST-2 mice (Shen et al., 1998). Although the FAST lines were bred in replicate using identical phenotyping procedures, due to their closed breeding populations and the independence of the two replicate lines, it is unlikely that they are genetically identical, either with regard to genes involved in the selection phenotype or background genes. In fact, this is evidenced by the stronger stimulant response of FAST-2 compared to FAST-1 mice, and the fact that FAST-1 mice responded more strongly to reverse selection (Phillips et al., 2002). This may suggest that additional mechanisms, absent in FAST-1 mice, influence the ethanol stimulant response in FAST-2 mice, mechanisms that may be under less control by GABAB receptors. This mechanism could be more exclusively involved at lower ethanol doses in this line. However, it is clear from the data in FAST-1 mice that the attenuation of the stimulant response to ethanol can occur in the absence of an accentuation of ataxia, an effect dramatically different from that seen with GABAA receptor agonists.
While accentuation of motor incoordination and sedation by baclofen was not seen in the current studies, previous studies have reported this. For example, high doses of baclofen potentiated ethanol-induced sedation in C57BL/6J mice (Besheer et al., 2004), and baclofen accentuated ethanol-induced motor incoordination both on the rotarod (Dar, 1996) and on a bar-holding task (Martz et al., 1983). It is possible that our results differ from others due to differences in genotype, dose, and procedure. However, enhancement of ethanol-induced sedation has typically been found with higher baclofen doses than those that were used here (e.g., 10 mg/kg or higher), doses that produce sedative effects on their own (Besheer et al., 2004). In addition, baclofen accentuated ethanol-induced ataxia using a fixed speed rotarod procedure set at 24 RPM (Dar, 1996), a considerably more difficult task than the 3 RPM requirement in the current study. While it is clear that motor incoordination and sedation are effects of GABAB receptor activation, our results suggest that the severity of these effects, at doses that attenuate ethanol’s stimulant effects, is dramatically less than that of GABAA receptor agonists.
In summary, activation of GABA systems attenuates acute behavioral sensitivity to ethanol; however, GABAB receptor specific drugs may produce fewer unwanted side effects than GABAA receptor specific drugs when given in combination with ethanol. Drugs that activate GABAA receptors act additively with ethanol and shift the behavioral response to ethanol towards greater intoxication, whereas drugs that activate GABAB receptors appear to reduce ethanol sensitivity via a more selective attenuation of the neurobiological effects of ethanol. These results have significant implications in the consideration of potential pharmacotherapies for alcohol use disorders. Reductions in ethanol consumption, such as seen with the GAT1 inhibitor tiagabine and other GABAA receptor acting drugs, may be due to a significant enhancement of ethanol’s negative side effects, including motor incoordination and sedation, rather than a reduction in reinforcement. Problematically, tolerance may develop to these side effects, resulting in a return to baseline drinking (see Nguyen et al., 2005). GABAB receptor agonists, at carefully titrated doses that minimize side effects, may hold greater promise as treatments for alcohol use disorders due to their reduced negative interaction effects with ethanol.
This work was supported by a grant from the Department of Veterans Affairs (TJP) and the National Institute on Alcohol Abuse and Alcoholism, grants P60AA10760 (TJP) and F31AA016031 (SEH). The authors gratefully acknowledge the advice on experimental design from Dr. John Crabbe and Andy Cameron, statistical advice from Dr. John Belknap, and the expert technical assistance provided by Carrie McKinnon, Na Li, Helen Kamens, Angela Scibelli, Sue Burkhart-Kasch, and Lauren Brown.