Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Chronic administration of ethanol (EtOH) leads to the development of tolerance, or reduced sensitivity, to most of EtOH's pharmacological effects (Kalant et al., 1971). The efficacy of -aminobutyric acid (GABA), benzodiazepines, and barbiturates also is reduced in animals following exposure to chronic EtOH. These findings suggest that chronic EtOH administration produces tolerance to EtOH and cross-tolerance to other positive modulators of GABA_A receptors (Buck and Harris, 1990; Morrow, 1995; Khanna et al., 1997).

The progesterone metabolite 3-hydroxy-5-pregnan-20-one (3,5-P or allopregnanolone) is a recently identified potent positive modulator of GABA_A receptors (for review, see Paul and Purdy, 1992; Lambert et al., 1995; Gasior et al., 1999). Consistent with the cross-tolerance to benzodiazepine and barbiturate effects just described, recent work found that chronic EtOH exposure produced reduced sensitivity to the anticonvulsant effect of diazepam in male and female rats during EtOH withdrawal (Devaud et al., 1996, 1998). However, in similarly treated rats, enhanced sensitivity to the anticonvulsant effect of 3,5-P was observed during EtOH withdrawal (Devaud et al., 1996, 1998). Progressive increases in response to chronically administered drugs are termed sensitization, and characterize certain effects of EtOH, such as low-dose locomotor activation (Masur and Boerngen, 1980; Phillips et al., 1998). Because chronic administration of EtOH can produce divergent changes in sensitivity to a subsequent dose of EtOH (i.e., sensitization or tolerance, depending on the pharmacological effect; Khanna et al., 1997; Phillips et al., 1998), the observed sensitization to the anticonvulsant effect of 3,5-P during EtOH withdrawal may not generalize to other pharmacological effects of 3,5-P.

Recently, we have shown that C57BL/6J (B6) and DBA/2J (D2) mice differ in behavioral sensitivity to 3,5-P (Finn et al., 1997). B6 mice were more sensitive than D2 animals to the anxiolytic, locomotor stimulant, and anticonvulsant effects of 3,5-P. In contrast, D2 were more sensitive than B6 to the muscle relaxation and ataxia produced by 3,5-P. The B6 and D2 inbred strains differ markedly in basal seizure susceptibility (D2 > B6) as well as in a number of EtOH-related behaviors (Phillips and Crabbe, 1991). For example, D2 mice exhibit more severe handling-induced convulsions than B6 after withdrawal from both acute (Roberts et al., 1992) and chronic (Crabbe et al., 1983; Crabbe, 1998) EtOH administration. Therefore, the strain difference in sensitivity to the anticonvulsant effect of 3,5-P (i.e., B6 > D2) is consistent with their differences in basal seizure susceptibility and EtOH withdrawal severity (i.e., B6 < D2). That is, the greater sensitivity to 3,5-P in the B6 inbred strain might confer some protection against basal seizure susceptibility as well as EtOH withdrawal severity because 3,5-P is a potent positive modulator of GABA_A receptors.

Genetic differences in EtOH withdrawal hyperexcitability may be due in part to modulatory effects of endogenous GABA agonist steroids, such as 3,5-P. Such genetic differences in the modulatory effects of endogenous 3,5-P at GABA_A receptors could result from differences in sensitivity to 3,5-P, or from altered biosynthesis, following exposure to chronic EtOH; these differences could lead to changes in neural excitability during EtOH withdrawal. Recent findings in our laboratory in mice selectively bred for sensitivity (Withdrawal Seizure-Prone, WSP) and resistance (Withdrawal Seizure-Resistant, WSR) to handling-induced convulsions following exposure to chronic EtOH inhalation are consistent with this hypothesis. Briefly, WSP mice were cross-tolerant to the anticonvulsant effect of a single dose of 3,5-P, whereas sensitivity was unchanged in similarly treated WSR animals (D.A.F., unpublished data). If the change in sensitivity to the anticonvulsant effect following exogenous administration of 3,5-P is indicative of a change in sensitivity to endogenous 3,5-P concentration, then exposure to chronic EtOH may render an EtOH withdrawal sensitive strain (i.e., WSP or D2) cross-tolerant to 3,5-P, which would produce less of a positive modulatory effect at GABA_A receptors and therefore, would increase neuronal excitability.

Because B6 and D2 differ in behavioral sensitivity to 3,5-P (Finn et al., 1997) and chronic EtOH exposure produced enhanced sensitivity to the anticonvulsant effect of 3,5-P in rats (Devaud et al., 1996), we wanted to determine 1) whether chronic EtOH exposure would produce enhanced sensitivity to other pharmacological effects of 3,5-P and 2) whether the two genotypes would differ in the change in sensitivity to 3,5-P during EtOH withdrawal. Because B6 have mild, and D2 have severe EtOH withdrawal, we hypothesized that B6 would become more sensitive to the anticonvulsant effect of 3,5-P, whereas D2 would not. If this were the case (i.e., sensitization in B6 and cross-tolerance in D2), these differential changes in sensitivity to the anticonvulsant effect of 3,5-P between B6 and D2 would be consistent with their differences in EtOH withdrawal severity.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Subjects

Drug-naive male B6 and D2 mice were used in all experiments. The animals were purchased from the Jackson Laboratory (Bar Harbor, ME) at 6 weeks of age, housed four per cage with ad libitum access to food and water, and acclimated to a 12:12 h light/dark cycle for a minimum of 2 weeks before experimentation. All procedures adhere to the U.S. Public Health Service-National Institutes of Health Guidelines for the care and use of laboratory animals and were approved by two local Institutional Animal Care and Use Committees.

Chronic EtOH Administration

Animals were exposed to EtOH vapor or air for 72 h using our standardized method for inducing EtOH dependence (for details, see Terdal and Crabbe, 1994). During the experiment, the animals were housed in stainless steel 1/4-in. hardware cloth cages inside a large Plexiglas chamber. Three such cages, measuring 48 × 35 × 18 cm, could be suspended by a flat metal guide rail inside a 71 × 66.5 × 114-cm chamber made of 1/2-in. Plexiglas, with a separate door allowing access to each wire mesh cage. Each wire mesh cage was subdivided into six compartments, so that each large Plexiglas chamber could house a maximum of 18 individual cages of mice. Ten centimeters below each wire mesh cage was a stainless steel drop pan with absorbent underpads, which were changed daily. A closed food hopper and water bottle were freely available. Chamber temperatures ranged from 28 to 30°C, depending on the total number of mice in the chamber. Airflow rate in the large Plexiglas chambers was 55 l/min. Because the volume of each large chamber was 538 liters, the air or EtOH vapor in each chamber was replaced approximately every 10 min. The presence of three large Plexiglas chambers allowed the simultaneous treatment of EtOH- and air-exposed animals.

The present studies used the alcohol dehydrogenase inhibitor pyrazole hydrochloride (pyrazole; Sigma Chemical Co., St. Louis, MO) to stabilize blood EtOH concentration (BEC). On day 1, mice in the EtOH groups were weighed, injected i.p. with a priming dose of EtOH (1.5 g/kg) and pyrazole (1.0 mmol/kg), and exposed to EtOH vapor (6 mg EtOH/l air for D2 and 9 mg EtOH/l air for B6) inside inhalation chambers. Different vapor doses were used to achieve equal BECs for the genotypes so that any genetic differences in behavioral responses could not be ascribed to differences in EtOH pharmacokinetics. At 24 and 48 h, the animals were briefly removed from the chambers, weighed, injected i.p. with pyrazole, and placed back into the chamber. Tail blood samples were taken from a subset of the animals each day to monitor BECs. Air-exposed animals also received daily pyrazole injections, but were injected with saline on day 1 and were exposed to air inside the inhalation chamber. At 72 h, all animals were removed from the chambers and tail blood samples were taken for subsequent analysis of BEC. Tails were nicked for the air-exposed groups, but no blood sample was taken. The mice were housed in polypropylene cages with cob bedding, taken to a procedure room for behavioral testing, and weighed. At peak withdrawal (i.e., 5.5-9.5 h postremoval from the inhalation chamber) the EtOH- and air-exposed animals were tested for behavioral sensitivity to 3,5-P. Subsequent 3,5-P dose groups for each treatment and genotype were counterbalanced across the 4-h time frame for behavioral testing.

BEC Determination

A 20-µl sample of blood from the tip of the tail was added to 50 µl of chilled 5% ZnSO₄ and stored on ice. Fifty microliters of 0.3 N Ba(OH)₂ and 300 µl of distilled water were added to each sample. The samples were shaken for 5 s and centrifuged for 5 min at 12,000 rpm. The supernatant was transferred to crimp-top glass vials and analyzed for EtOH concentration by gas chromatography. Four pairs of external standards of known EtOH concentration (0.5-4.0 mg/ml) were used to establish a standard curve.

Behavioral Sensitivity to 3,5-P

Behavioral testing occurred between 1:00 PM and 5:00 PM (i.e., during peak EtOH withdrawal). Each mouse was tested at 10-min intervals on several behavioral tasks (Table 1). The procedures have been described recently (Finn et al., 1997) and will be outlined briefly. 1) Muscle relaxation. Mice are placed on a tray, allowed to grasp a horizontal bar connected to a strain gauge, and then pulled gently until they lose their grip. This procedure allows each mouse to be tested for baseline ability and ability following 3,5-P injection. The steroid effect was expressed as a change from baseline. 2) Ataxia. Each mouse was placed on a stationary Rotarod that began to accelerate linearly (20 rpm/min) in its rate of rotation until the mouse fell off. The latency to fall was then used to calculate the speed (rpm) at which the mouse could no longer remain on the Rotarod. This procedure allowed each mouse to be tested for Rotarod performance at baseline and following 3,5-P injection. The steroid effect was expressed as a change from baseline rpm. 3) Anxiolysis. The elevated plus maze consists of two open and two enclosed horizontal perpendicular arms extending from a central platform (5 × 5 cm), 50 cm above the floor. Each mouse was placed on the central platform facing an intersection between open and closed arms and allowed to explore freely for 5 min. During the 5-min test period, the number of entries into the open and closed arms as well as the amount of time spent in the open and closed arms was measured. For an arm entry to be measured, all four paws had to be within the arm. The percentages of open-arm time and open-arm entries were used as indices of anxiety. To demonstrate the predicted anxiogenic effect of EtOH withdrawal, we wanted to use conditions under which naïve animals spent equal time in both open and closed arms. Pilot testing showed that we could achieve this condition if we adjusted the sides of the open arm upward (from the initial configuration of 0.5 cm to 1.2-cm height; data not shown). 4) Activity. The information obtained from elevated plus maze testing (i.e., total number of arm entries) was used as an estimate of locomotor activity. 5) Seizure protection. Mice were administered the convulsant pentylenetetrazol (PTZ; 5 mg/ml in saline) via timed tail vein infusion into a lateral vein (0.5-ml/min infusion rate). The apparatus and procedure for tail vein infusion have been described in detail (Kosobud and Crabbe, 1990). Briefly, an animal was placed into a Plexiglas container (5-cm diameter) that allowed it to be hand held by the tail and visualized during the infusion. A 27-gauge butterfly needle, connected to the PTZ-containing syringe on an infusion pump (Sage Instruments, model 355; Orion, Boston, MA), was inserted into a lateral vein. The needle was held in place with one hand while the latencies for onset to myoclonic twitch (MC twitch), face and forelimb clonus (FF clonus), running bouncing clonus (RB clonus), and tonic hindlimb extension (THE) were recorded in seconds. Subsequently, the latencies were converted to threshold convulsant dosage of PTZ to elicit each convulsion endpoint (i.e., milligram of drug per kilogram body weight). Therefore, this method allowed for observation and qualitative analysis of several different endpoints that characterize PTZ-induced convulsions (i.e., MC twitch, FF clonus, RB clonus, and THE).

View this table: [in this window] [in a new window]   TABLE 1 Sequence of behavioral tests The infusion was terminated once the animals had exhibited THE or at 4 min (in cases where the mice were protected by the dose of 3,5-P). The animal was euthanized and trunk blood was collected for subsequent analysis of plasma corticosterone and 3,5-P concentration by radioimmunoassay.

Radioimmunoassay (RIA)

On completion of behavioral testing, the mice were euthanized and trunk blood collected for subsequent analysis of plasma 3,5-P and corticosterone (CORT) concentration by RIA.

3,5-P. The RIA for 3,5-P was adapted from the method of Purdy et al. (1990) and is described in detail elsewhere (Finn and Gee, 1994). The RIA used [³H]3,5-P (54 Ci/mmol; New England Nuclear, Boston, MA) and a polyclonal antiserum that was a generous gift from CoCensys (Irvine, CA). Counts per minute were normalized and fit to a least-squares regression equation produced by log-logit transformation of the standards. Mass of samples was calculated by interpolation of the standards and correction for recovery. The minimum detectable limit in the present assay was 25 pg. The intra-assay coefficient of variation averaged 14%, and the interassay coefficient of variation in seven assays averaged 15%.

CORT. Plasma (5 µl) was diluted with 100 µl of sterile water and stored at 4°C until assayed. Samples were immersed in boiling water for 5 min to denature CORT-binding globulin. The RIA was adapted from a previously reported procedure (Keith et al., 1978) and used ¹²⁵I-CORT from ICN Pharmaceuticals (Costa Mesa, CA) and antisera from Ventrex (Portland, ME). Counts per minute were normalized and fit to a least-squares regression equation produced by log-logit transformation of the standards. Mass of samples was calculated by interpolation of the standards. The detectable range of the assay was from 0.1 to 400 µg of CORT per 100 ml of plasma. Intra- and interassay coefficients of variation were <10%. The specificity of the assay is very high, with only 4% cross-reactivity to deoxycorticosterone, 1% cross-reactivity to 5-pregnanedione, and <0.6% cross-reactivity to other endogenous steroids.

Data Analysis

Data are expressed as the means ± S.E. ANOVA was used to assess strain and dose effects on BEC, to ensure that the two genotypes were matched for EtOH exposure. Then ANOVA was used to determine the influence of strain, treatment, and dose on the dependent variables muscle relaxation (percentage of baseline forelimb grip strength), ataxia (percentage of baseline Rotarod performance), activity (total arm entries on elevated plus maze), anxiolysis (percentage open-arm time and open-arm entries on plus maze), seizure protection (threshold dose for onset to MC twitch, FF clonus, RB clonus, and THE), plasma 3,5-P concentration, and plasma CORT concentration. Due to differences in susceptibility to PTZ in the EtOH- versus air-exposed mice, the seizure protection data were transformed to percentage of change in PTZ threshold dose of the mean for the respective vehicle-injected group. If three-way interactions were obtained, the data for each strain were then analyzed separately. When appropriate, simple main effects analyses followed by Tukey post hoc comparisons were then used to examine significant treatment and dose effects within each strain. Mice with less than two total entries on the elevated plus maze were eliminated from the plus maze analysis (n = 6 EtOH-exposed B6, n = 5 EtOH-exposed D2, and n = 1 air-exposed B6). In addition, data from four air-exposed B6 animals with incorrect injections were eliminated from the analyses. Significance was set at P  .05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Exposure to EtOH vapor produced stable BEC at the time of removal from the inhalation chamber. Mean ± S.E. BEC was 1.05 ± 0.06 mg/ml for B6 and 1.21 ± 0.05 mg/ml for D2. There was a trend for a difference between the strains in BEC (F = 3.83; df = 1, 79; P = .054). However, there was no effect of 3,5-P dose or interaction between strain and dose (both F < .8). These results indicate that the chronic EtOH exposure was similar in the two genotypes when they subsequently were divided into different 3,5-P dose groups.

Elevated Plus Maze. Total number of entries on the elevated plus maze, which was used as an index of activity, was significantly decreased during EtOH withdrawal in both B6 and D2 mice (Fig. 1). This conclusion is supported by the significant main effect of treatment (F = 115.86; df = 1, 140; P = .0001). Although there was no overall main effect of strain on total entries, there was a significant effect of dose of 3,5-P (F = 5.04; df = 3, 140; P < .003) on the total number of entries on the elevated plus maze. Interactions between strain and treatment (F = 9.93; df = 1, 140; P < .003) and between strain and dose (F = 12.31; df = 3, 140; P < .0001) were significant, but the three-way interaction was not significant.

View larger version (10K): [in this window] [in a new window]   Fig. 1.   The effect of 3,5-P on total entries, an index of activity, in air- and EtOH-exposed B6 (A) and D2 (B) mice. Animals were tested on the elevated plus maze at 10-min postinjection of 3,5-P or vehicle at a time corresponding to peak withdrawal in the EtOH-exposed mice. Values represent means ± S.E. The n/dose of 3,5-P for each treatment and genotype was as follows. For B6, n = 10/dose (vehicle and 3.2 mg/kg) or 7/dose (10 and 17 mg/kg) for EtOH-exposed mice and n = 9/dose (vehicle and 3.2 mg/kg), n = 8 (10 mg/kg) or n = 7 (17 mg/kg) for air-exposed mice. For D2, n = 10/dose (3.2 and 10 mg/kg) or n = 8 (vehicle) or n = 6 (17 mg/kg) for the EtOH-exposed D2 and n = 10/dose for the air-exposed D2.

View larger version (11K): [in this window] [in a new window]   Fig. 2.   Sensitivity to the anxiolytic effect of 3,5-P in air- and EtOH-exposed B6 (A) and D2 (B) mice, measured by the percentage of time spent in the open arms of the elevated plus maze. Values represent means ± S.E. for the animals depicted in Fig. 1. *P < .05 versus respective vehicle-treated group; +P < .05 versus respective EtOH-exposed dose group of 3,5-P.

View larger version (11K): [in this window] [in a new window]   Fig. 3.   Sensitivity to the anxiolytic effect of 3,5-P in air- and EtOH-exposed B6 (A) and D2 (B) mice, measured by the percentage of open-arm entries on the elevated plus maze. Values represent means ± S.E. for the animals depicted in Fig. 1.

Forelimb Grip Strength. A decrease in percentage of baseline forelimb grip strength following injection was used as an indication of muscle relaxation. The percentage of baseline grip strength was significantly influenced by treatment (F = 13.78; df = 1, 152; P < .0005) and dose of 3,5-P (F = 33.33; df = 3, 152; P < .0001), with a trend for a main effect of strain (F = 3.64; df = 1, 152; P = .058) (Fig. 4). Interactions between strain and dose (F = 3.15; df = 3, 152; P < .03) and between treatment and dose (F = 2.99; df = 3, 152; P < .04) were significant, with a trend for a three-way interaction between main effects (F = 2.49; df = 3, 152; P = .06). When the analysis was limited to the animals injected with vehicle, there was no effect of treatment on percentage of baseline grip strength, suggesting that EtOH withdrawal did not influence baseline grip strength in either B6 or D2 mice.

View larger version (11K): [in this window] [in a new window]   Fig. 4.   Differences in sensitivity to 3,5-P-induced muscle relaxation in B6 (A) and D2 (B) mice exposed to air or EtOH vapor. Shown is the mean ± S.E. percentage of baseline forelimb grip strength in animals that were tested 20 min before, and 20 min post, injection of 3,5-P or vehicle. The n/dose was 10 for the EtOH- and air-exposed D2. There were 10/dose for the EtOH-exposed B6 and either 8/dose (vehicle), 9/dose (3.2 and 10 mg/kg) or 7/dose (17 mg/kg) for the air-exposed B6. *P < .05, **P < .01, ***P < .001 versus respective vehicle-treated group; +P < .05, ++P < .01 versus respective EtOH-exposed dose group of 3,5-P.

Rotarod Performance. A decrease in percentage of baseline Rotarod performance was used as an indication of ataxia. The percentage of baseline Rotarod performance was significantly influenced by dose of 3,5-P (F = 8.96; df = 3, 150; P < .0001), with a trend for a difference between the two genotypes (F = 3.41; df = 1, 150; P = .067) (Fig. 5). Although there was no significant effect of treatment on percentage of baseline Rotarod performance, there was a trend for an interaction between strain and treatment (F = 3.09; df = 1, 150; P = .08) and between treatment and dose of 3,5-P (F = 2.50; df = 3, 150; P = .06). Simple main effects analysis indicated that B6 and D2 mice differed in percentage of baseline Rotarod performance following exposure to chronic EtOH (F = 5.26; df = 1, 150; P < .03) (i.e., B6 > D2), but not following exposure to air. Analyses subsequent to the treatment by dose interaction indicated that percentage of baseline Rotarod performance was significantly greater in the EtOH- versus air-exposed mice injected with the 17-mg/kg dose of 3,5-P (F = 5.33; df = 1, 150; P < .03), an effect that appeared to be due primarily to the results in B6 mice. In addition, the effect of 3,5-P dose on percentage of baseline Rotarod performance was significant in the air-exposed mice (F = 7.45; df = 3, 150; P = .0001), with a trend for significance in the EtOH-exposed animals (F = 2.63; df = 3, 150; P = .052). Although these results are not conclusive, they are suggestive that the ataxic effect of 3,5-P, measured by a significant decrease in percentage of baseline Rotarod performance, was more pronounced in air- versus EtOH-exposed B6 mice.

View larger version (11K): [in this window] [in a new window]   Fig. 5.   Sensitivity to the ataxic effect of 3,5-P in air- and EtOH-exposed B6 (A) and D2 (B) animals. Rotarod performance was tested at 10 min before, and 30 min post, injection of 3,5-P or vehicle. Values represent means ± S.E. percentage of baseline Rotarod performance for the animals depicted in Fig. 4, with the exception that the n = 9 for the EtOH-exposed B6 injected with 10-mg/kg 3,5-P and for the EtOH-exposed D2 injected with 17-mg/kg 3,5-P. +P < .05 versus respective EtOH-exposed dose group of 3,5-P, collapsed across genotype.

Seizure Susceptibility. Drugs that alter seizure threshold can be tested for pro- or anticonvulsant activity by pretreating mice and observing the effect on PTZ seizure threshold. A decreased PTZ seizure threshold indicates proconvulsant activity, whereas an increased PTZ threshold suggests anticonvulsant activity. By administering PTZ via timed tail vein infusion, distinct convulsive responses were produced as a function of dose. Although these seizure manifestations might appear to be on a continuum, there are two qualitatively distinct seizure components that are mediated by separable and independent anatomical circuits (Gale, 1988). MC twitch and FF clonus appear to be associated with forebrain neural circuits, whereas RB clonus and THE depend on hindbrain circuitry. Genetic susceptibility to these two seizure types also may be distinct (Kosobud and Crabbe, 1990). Therefore, results for MC twitch and FF clonus were analyzed as similar types of convulsions as were results for RB clonus and THE.

View larger version (14K): [in this window] [in a new window]   Fig. 6.   Differential basal sensitivity to PTZ in EtOH- and air-exposed B6 (A) and D2 (B) mice injected with vehicle. PTZ was administered via timed tail vein infusion at 40-min postinjection. Values represent means ± S.E. for 10 mice per strain and treatment, with the exception of the air-exposed B6 where the n = 9. *P < .05, **P < .01, ***P < .001 versus respective air-exposed group.

View larger version (11K): [in this window] [in a new window]   Fig. 7.   Differential change in sensitivity to the anticonvulsant effect of 3,5-P, measured by the percentage of change in threshold dose for onset to MC twitch in B6 (A) and D2 (B) mice exposed to chronic EtOH vapor or air. PTZ was administered at 40-min postinjection of 3,5-P or vehicle. Values represent means ± S.E. for 9-10/dose with the exception of the air-exposed B6 injected with 17-mg/kg where the n = 7. *P < .05, **P < .01, ***P < .001 versus respective vehicle-treated group; +++P < .001 versus respective EtOH-exposed dose group of 3,5-P

View larger version (11K): [in this window] [in a new window]   Fig. 8.   Sensitivity to the anticonvulsant effect of 3,5-P, measured by the percentage of change in threshold dose for onset to THE in B6 (A) and D2 (B) mice. Shown are the means ± S.E. for the mice depicted in Fig. 7. **P < .01, ***P < .001 versus respective vehicle-treated group; +++P < .001 versus respective EtOH-exposed dose group of 3,5-P.

RIA. Plasma 3,5-P concentration in the EtOH- and air-exposed B6 and D2 mice following completion of the behavioral testing (i.e., ~50 min postinjection of 3,5-P or vehicle) is depicted in Fig. 9. Dose of 3,5-P (F = 77.505; df = 3, 149; P < .0001) significantly influenced plasma 3,5-P concentration, producing a dose-dependent increase in plasma 3,5-P levels in both genotypes. There was a trend for an effect of strain on plasma 3,5-P concentration (F = 2.81; df = 1, 149; P < .10), whereas the effect of treatment was not significant. In addition, there was no significant interaction between strain, treatment and dose of 3,5-P, suggesting that any change in sensitivity to 3,5-P during EtOH withdrawal was not due to treatment differences in plasma 3,5-P concentration among genotypes.

View larger version (17K): [in this window] [in a new window]   Fig. 9.   Plasma 3,5-P concentration in EtOH-exposed and air-exposed B6 (A) and D2 (B) mice following completion of the behavioral testing (i.e., ~50 min after injection of 3,5-P). Basal 3,5-P levels are depicted in (C). Values represent means ± S.E. for the animals depicted in Figs. 1-8.

View larger version (18K): [in this window] [in a new window]   Fig. 10.   Plasma CORT levels in EtOH- and air-exposed B6 (A) and D2 (B) mice following completion of the behavioral testing. Basal CORT concentration is depicted in (C). Values represent means ± S.E. for the animals depicted in Figs. 1-9.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Chronic EtOH exposure produced changes in sensitivity to exogenous administration of a GABA agonist steroid, which varied, depending on the genotype and pharmacological effect. Recent findings that chronic EtOH exposure enhanced neuroactive steroid stimulation of [³H]muscimol binding (Negro et al., 1993) as well as potentiation of GABA_A receptor-mediated chloride influx (Devaud et al., 1996) suggested that the sensitization to the anticonvulsant effect of 3,5-P that was demonstrated in rats (Devaud et al., 1996, 1998) might generalize to all of the pharmacological effects of 3,5-P. Consistent with the results in rats, the present findings indicate that sensitization to the anticonvulsant effect of 3,5-P was observed in B6 mice during EtOH withdrawal. In contrast, EtOH-exposed D2 mice were either cross-tolerant or exhibited no change in sensitivity to the anticonvulsant effect of 3,5-P, depending on the convulsion endpoint. The only other clear change in sensitivity to 3,5-P with chronic EtOH exposure occurred with muscle relaxation, where sensitivity to the muscle relaxant effect of 3,5-P was reduced during EtOH withdrawal in both genotypes. Importantly, these differences in behavioral sensitivity to 3,5-P in EtOH- versus air-exposed B6 and D2 mice were not due to differences between groups in plasma 3,5-P concentration. Therefore, in conjunction with the recent report that rats were sensitized to the anticonvulsant effect of 3,5-P during EtOH withdrawal (Devaud et al., 1996, 1998), the present findings suggest that sensitization to the anticonvulsant effect of 3,5-P during EtOH withdrawal does not generalize across all genotypes nor does it generalize to all of the pharmacological effects of 3,5-P.

The differential change in sensitivity to the anticonvulsant effect of 3,5-P during EtOH withdrawal in B6 versus D2 mice was not due to strain differences in EtOH exposure because there were no differences in the BEC among the groups of animals that were injected with vehicle or 3,5-P before behavioral testing during EtOH withdrawal. In the present study, BEC was determined immediately on initiation of withdrawal (i.e., on removal from inhalation chambers) and was not determined immediately before the behavioral testing. However, separate studies (D.A.F., unpublished data) have determined that BEC at 6-h postremoval from the inhalation chambers was negligible (mean ± S.E., 0.06 ± 0.03 mg/ml; n = 13) in animals with an initial BEC similar to that in the present study (1.16 ± 0.05 mg/ml). Therefore, it is unlikely that strain differences in EtOH pharmacokinetics contributed to the present findings as blood EtOH should have been nearly eliminated at the time of behavioral testing.

Chronic EtOH exposure produces well documented bidirectional changes in GABA_A receptor subunit gene expression (Buck et al., 1991; Montpied et al., 1991; Mhatre et al., 1993; Mhatre and Ticku, 1994; Devaud et al., 1995, 1996; Mahmoudi et al., 1997). Overall, 1- and 2-subunit mRNA levels significantly decrease, whereas 4-, 1-3-, 1-, and 2S-subunit mRNA levels were significantly increased following various chronic EtOH exposure paradigms. Comparison between the EtOH-induced changes in GABA_A receptor subunit mRNA levels versus peptide levels suggests that they are highly correlated during EtOH dependence (Devaud et al., 1997). Because studies using recombinantly expressed receptors have demonstrated that the GABA_A receptor subunit composition can determine the pharmacological properties of these receptors (Sieghart, 1995), it is possible that chronic EtOH-induced changes in GABA_A receptor subunit mRNA and peptide levels would lead to alterations in assembly of receptors and in sensitivity of these receptors to positive modulators.

With regard to neuroactive steroids, potentiation of GABA_A receptor function does not exhibit an absolute requirement for any specific subunit (Lambert et al., 1995). However, recent work has found that recombinantly expressed receptors incorporating the 1- (Puia et al., 1993) or 6- (Im et al., 1994; Hauser et al., 1995; Lambert et al., 1996) subunits had enhanced sensitivity to neuroactive steroids, whereas inclusion of the -subunit (Zhu et al., 1996) or recently identified -subunit (Davies et al., 1997) reduced sensitivity of these receptors to neuroactive steroids. Therefore, it is possible that strain and species differences in the brain regional distribution of GABA_A receptors as well as in the chronic EtOH-induced alterations in subunit composition of GABA_A receptors or post-translational modifications underlie the genotypic differences in neuroactive steroid sensitivity during EtOH withdrawal. The effects of chronic EtOH exposure on gene expression of specific GABA_A receptor subunits are currently unknown in B6 and D2 mice. These studies are underway (K. Buck, unpublished data) and hopefully will provide information on the potential role of different GABA_A receptor isoforms in the differential changes in sensitivity to 3,5-P during EtOH withdrawal.

Recent findings indicate that intra-amygdala administration of benzodiazepines produces anxiolysis without locomotor stimulation, muscle relaxation, ataxia, or seizure protection (Lotrich and Gallaher, 1998). Therefore, the bidirectional changes in sensitivity to the various pharmacological effects of 3,5-P in the present study suggest that chronic EtOH exposure is producing differential changes in GABA_A receptors in the various brain regions that may underlie the distinct pharmacological properties of 3,5-P. Importantly, the enhanced sensitivity to the anticonvulsant effect and reduced sensitivity to the muscle relaxant effect of 3,5-P in B6 mice emphasizes the therapeutic potential of neuroactive steroids or their analogs in the treatment of alcohol dependence.

Even though the B6 and D2 strains were matched for chronic EtOH exposure, the two genotypes differ markedly in withdrawal severity (i.e., D2 > B6), when it is indexed by an increase in handling-induced convulsions following removal from the inhalation chambers (Crabbe, 1998). It is noteworthy that the enhanced sensitivity to the anticonvulsant effect of 3,5-P was evident only in B6 mice following exposure to chronic EtOH, an inbred strain that has modest withdrawal compared with a panel of inbred strains (Crabbe et al., 1983; Crabbe, 1998). This finding is consistent with our prediction that the change in sensitivity to the anticonvulsant effect of 3,5-P would be inversely related to withdrawal severity (i.e.,  sensitivity,  withdrawal).

Withdrawal from chronic EtOH exposure also can produce changes in sensitivity to convulsant drugs (Watson and Little, 1995; Finn and Crabbe, 1999; Mhatre and Gonzalez, 1999). In the present study, basal seizure susceptibility to PTZ was increased in EtOH-withdrawing B6, but not D2, mice injected with vehicle. Although this finding might suggest that withdrawal severity in B6 was greater due to the significant increase in sensitivity to PTZ in the EtOH- versus air-exposed mice, EtOH naive D2 were much more sensitive to PTZ than were B6 mice (i.e., compare air-exposed B6 versus D2 in Fig. 6). Therefore, the lack of a significant change in sensitivity to PTZ in EtOH- versus air-exposed D2 mice may be due to a "floor effect" in that it would be difficult to detect a further decrease in PTZ threshold dose with the current experimental paradigm because this genotype is extremely sensitive to PTZ in the absence of EtOH (Kosobud and Crabbe, 1990). It is also possible that independent mechanisms contribute to PTZ-induced versus EtOH withdrawal-induced convulsions. Nonetheless, when EtOH withdrawal is indexed by convulsions elicited by handling, withdrawal severity is greater in the D2 inbred strain.

The predicted anxiogenic effect of EtOH withdrawal was not apparent in the present study. An elevated plus maze was used to measure anxiety because this animal model (Pellow et al., 1985; Lister, 1987) is able to detect both anxiolytic and anxiogenic drug effects in mice (Lister, 1987). Because naïve animals typically spend ~25% of the test time in the open arms, the sides of the open arms were adjusted upward so that naïve animals would spend equal time in both open and closed arms, to facilitate detection of an anxiogenic effect of EtOH withdrawal. Therefore, it was surprising that the vehicle-injected, air-exposed mice spent less than the predicted 50% of time in the open arm. In a separate study in which B6 and D2 mice were tested only on the elevated plus maze during peak withdrawal from exposure to 72 h of EtOH vapor or air, there was a significant decrease in open-arm entries in the EtOH-exposed D2, but not B6, mice versus their respective air-exposed controls (D.A.F., unpublished data). Therefore, we are able to detect an anxiogenic effect of EtOH withdrawal following exposure to 72 h of EtOH vapor, with a greater increase in anxiety in D2 versus B6 mice, when the animals are not tested on several behavioral tasks. Consequently, it is unclear if the pyrazole injections in the air-exposed animals or the baseline grip strength and Rotarod testing of the animals before injection in the present study produced an increased level of anxiety in the air-exposed mice, which then made it difficult to detect an anxiogenic effect of withdrawal in the EtOH-exposed animals. Consistent with this notion, plasma CORT concentrations were elevated in the air-exposed animals (i.e., > 20 µg/dl) compared with a typical basal plasma CORT concentration of 2 to 5 µg/dl in naïve animals or plasma CORT levels of 6 to 8 µg/dl in air-exposed B6 and D2 mice that were not behaviorally tested (D.A.F., unpublished data). This suggests that the air-exposed mice were moderately stressed.

It is unlikely that stress associated with repeated testing of an animal contributed to the change in sensitivity to 3,5-P that was observed following chronic EtOH exposure. First, we have previously validated the methodology for repeated testing of an animal and demonstrated that seizure susceptibility to PTZ, as well as sensitivity to the anticonvulsant effect of 3,5-P, was not different in B6 or D2 mice that were tested once versus repeatedly tested before the seizure susceptibility determination (Finn et al., 1997). Second, comparison of plasma 3,5-P and CORT concentrations in vehicle-injected animals in the present study with values from similarly treated B6 and D2 mice that were not behaviorally tested (D.A.F., unpublished data) suggests that plasma 3,5-P levels did not differ in tested versus untested mice (i.e., values in EtOH- and air-exposed B6 and D2 mice were similar to that in the present study). However, plasma CORT concentrations did differ in that values in air-exposed mice were significantly lower than those in the present study (as discussed above) and were significantly increased during EtOH withdrawal in both genotypes. Therefore, the lack of significant differences in basal 3,5-P concentrations in tested versus untested mice, coupled with the similar innate sensitivity to the anticonvulsant effect of 3,5-P in repeatedly tested versus singly tested EtOH naïve mice, suggests that stress associated with activation of the hypothalamic-pituitary-adrenal axis did not influence sensitivity to the pharmacological effects of 3,5-P in the present study.

In conclusion, the present results indicate that two inbred strains that differ markedly in chronic EtOH withdrawal severity also differ in the change in sensitivity to the anticonvulsant effect of 3,5-P during EtOH withdrawal. If the present findings with exogenous administration of 3,5-P are indicative of a change in sensitivity to endogenous 3,5-P concentration, then exposure to chronic EtOH may render an EtOH-withdrawal-resistant strain more sensitive to 3,5-P than an EtOH-withdrawal-sensitive strain. Importantly, this relationship between genetic differences in EtOH withdrawal severity and 3,5-P sensitivity only may be relevant to the anticonvulsant effect because the sensitization to the anticonvulsant effect of 3,5-P during EtOH withdrawal in B6 mice did not generalize to all pharmacological effects of 3,5-P.