logy 552 (2006) 83–89www.elsevier.com/locate/ejphar

European Journal of Pharmaco

The motor-impairing effects of GABAA and GABAB agonists inγ-hydroxybutyrate (GHB)-treated rats: Cross-tolerance

to baclofen but not flunitrazepam

Mark A. Smith a,b,⁎, Samantha R. Gergans a, Megan A. Lyle a,b

a Department of Psychology, Davidson College, Davidson, NC, USAb Program in Neuroscience, Davidson College, Davidson, NC, USA

Received 1 July 2006; received in revised form 28 August 2006; accepted 31 August 2006Available online 12 September 2006

Abstract

γ-Hydroxybutyrate (GHB) is believed to function as a neurotransmitter in the mammalian brain by binding to a GHB-specific binding site. Inaddition, GHB may also indirectly enhance the neuroinhibitory actions of γ-aminobutyric acid (GABA) by converting to GABA at neuronalsynapses. The purpose of the present study was to examine the effects of representative GABAA and GABAB receptor agonists in rats treatedchronically with GHB. Using a rotorod apparatus, the motor-impairing effects of GHB, the indirect GABAA receptor agonist, flunitrazepam, andthe direct GABAB receptor agonist, baclofen, were examined before, during and after chronic treatment with 1000 mg/kg GHB, b.i.d. Prior tochronic treatment, all three drugs produced dose-dependent decreases in motor performance at low (8 rpm) and high (32 rpm) rotational speeds.Chronic treatment with GHB significantly decreased the potency of baclofen at both speeds, but did not alter the potency of either GHB orflunitrazepam. Following termination of chronic treatment, the potency of baclofen increased significantly at both speeds and returned to thatobserved prior to chronic treatment. These data indicate that chronic treatment with GHB confers tolerance to a GABAB receptor agonist underconditions in which tolerance is not conferred to a GABAA receptor agonist. These findings are consistent with the in vivo behavioral profile ofGHB, which reveals a greater role for GABAB receptors than for GABAA receptors in its behavioral effects.© 2006 Elsevier B.V. All rights reserved.

Keywords: γ-hydroxybutyrate (GHB); Flunitrazepam; Baclofen; Cross-tolerance; Rotorod; Motor performance

1. Introduction

γ-Hydroxybutyrate (GHB) is a centrally acting agent that isclinically marketed as a treatment for narcolepsy and has beenused in Europe to clinically manage the withdrawal syndromeassociated with alcohol abuse and dependence (Kam and Yoong,1998; Tunnicliff and Raess, 2002). GHB has also been identifiedas an emerging drug of abuse, and was classified as a Schedule Idrug under the U.S. Controlled Substances Act in 2000 (DrugEnforcement Agency, 2000). GHB has also been implicated as adrug used to facilitate acquaintance rape (Schwartz et al., 2000),presumably because of its motor-impairing effects. Indeed,epidemiological studies have reported the presence of GHB and

⁎ Corresponding author. Department of Psychology, Davidson College,Davidson, NC 28035-7037, U.S.A. Tel.: +1 704 894 2470; fax: +1 704 894 2512.

E-mail address: masmith@davidson.edu (M.A. Smith).

0014-2999/$ - see front matter © 2006 Elsevier B.V. All rights reserved.doi:10.1016/j.ejphar.2006.08.080

its metabolites in urine samples taken from victims of allegedsexual assault, often in combination with other drugs such asethanol and flunitrazepam (ElSohly and Salamone, 1999).

GHB has also been identified as a putative neurotransmitterin the mammalian central nervous system. Data obtained invitro indicate that GHB is synthesized in neurons, released atpresynaptic nerve terminals in a Ca2+-dependent manner, andtaken back into the presynaptic terminal by a specific reuptakemechanism (Maitre, 1997). GHB binds to a high-affinitybinding site that that is believed to be a GHB-specific receptorwidely distributed in mammalian brain and distinct and separatefrom other known neurotransmitter receptors (Mathivet et al.,1997; Lingenhoehl et al., 1999). Evidence implicating a GHB-specific receptor as a primary mechanism of action for GHB ismixed, as the putative GHB-receptor antagonist, NCS-382, isonly partially effective at blocking the physiological andbehavioral effects produced by GHB (Cook et al., 2002).

84 M.A. Smith et al. / European Journal of Pharmacology 552 (2006) 83–89

GHB also serves as a precursor molecule for γ-aminobutyricacid (GABA), and there is increasing evidence that many of itsactions may be mediated by GABA-specific receptors. GHBbinds to a low-affinity binding site, which is believed to be aGABAB receptor, but evidence that GHB functions as a directGABAB agonist is mixed (Feigenbaum and Howard, 1996).Several researchers have postulated that GHB functions as anindirect agonist at both GABAA and GABAB receptors byconverting to GABA in the central nervous system (Della Pietraet al., 1966; Vayer et al., 1985; Hechler et al., 1997). Consistentwith this hypothesis, GHB shares discriminative stimulus effectswith the indirect GABAA receptor agonist, diazepam (Colomboet al., 1998; Carter et al., 2003), and the direct GABAB receptoragonist, baclofen (Colombo et al., 1998; Lobina et al., 1999).Evidence for the role of GABAB receptors in the behavioraleffects of GHB is also observed in studies showing that theselective GABAB receptor antagonist, CGP 35348, antagonizesthe discriminative stimulus effects of GHB in GHB-trainedanimals (Carter et al., 2003).

Studies of tolerance and cross-tolerance are often used todetermine whether two or more compounds share a commonmechanism of action. For instance, when administered chron-ically, indirect GABAA agonists produce cross-tolerance toother indirect GABAA agonists (Le et al., 1986; Byrnes et al.,1993), but do not produce cross-tolerance to other drugs,including agonists at GABAB receptors (Chodera et al., 1984;Xie and Tietz, 1992). Similarly, chronic administration of aGABAB agonist produces tolerance to its effects (Hefferan et al.,2006), but does not produce cross-tolerance to the effects ofother drugs (Levy and Proudfit, 1977). Recently, Eckermannet al. (2004) reported that chronic administration of the GHBprecursor, 1,4-butanediol, produces cross-tolerance to bothGHB and baclofen, suggesting that these drugs may share acommon mechanism of action. Although a few studies haveexamined the development of tolerance to GHB in GHB-treatedanimals (e.g., Bania et al., 2003; Itzhak and Ali, 2002); thosestudies typically have not examined the development of cross-tolerance to other drugs (but see Colombo et al., 1995).

The purpose of the present study was to examine the motor-impairing effects of representative agonists at GABAA andGABAB receptors in rats treated chronically with GHB. To thisend, eight male rats were trained to walk on a rotorod apparatusrotating at either low (8 rpm) or high (32 rpm) speeds, and testedwith the indirect GABAA receptor agonist, flunitrazepam, andthe direct GABAB receptor agonist, baclofen, before, during andafter chronic treatment with 1000 mg/kg GHB, b.i.d. The effectsof each drug were examined at multiple speeds because previousstudies have shown that the potency and effectiveness of GABA-receptor agonists in this procedure depend on the difficulty of thetask (Smith and Stoops, 2001; Smith et al., 2004).

2. Materials and method

2.1. Animals

Eight male, Long-Evans rats, each weighing approximately280 g upon arrival, were obtained from Charles River

Laboratories (Raleigh, NC, USA). All subjects were housedindividually in a temperature-controlled colony room main-tained on a 12-h light/dark cycle (lights on: 07:00) with watercontinuously available in the home cage. In order to holdpharmacokinetic variables related to weight and body massconstant, rats were maintained at 300–350 g over the course ofbehavioral testing via light food restriction. One rat died forreasons unrelated to drug administration prior to chronictreatment with GHB, and only data from the seven rats thatcompleted the study are presented. All subjects were tested andmaintained in accordance with the guidelines of the InstitutionalAnimal Care and Use Committee of Davidson College and theGuide for the Care and Use of Laboratory Animals (Institute ofLaboratory Animal Resources, 1996).

2.2. Apparatus

Motor performance was measured with the use of a rotorodapparatus from San Diego Instruments, Inc. (San Diego, CA,USA). The rotorod consisted of a steel and Plexiglas cabinet(48 cm×50 cm×102 cm) that was divided into four 12-cm lanesby vertical partitions. A 10-cm rubber cylinder was suspended80 cm above the floor of the apparatus and ran through the centerof each lane. Rotation of the cylinder was controlled by theoperator via a motorized drive unit that controlled the speed ofrotation from 0 to 101.3 rpm in 0.1-rpm increments. The latencyto fall from the apparatus was recorded by the interruption of aphotobeam located 12 cm below the cylinder. Towels wereplaced on the floor of the apparatus to cushion the impact of allfalls.

2.3. Behavioral training

Subjects were trained and tested according to methodsdescribed previously (Smith and Stoops, 2001; Smith et al.,2004). Briefly, rats were trained to walk on the rotorodapparatus during daily training sessions in which the speed ofthe apparatus was increased from 3 rpm to 32 rpm in 2- to 3-rpmincrements over the course of 12 days. Throughout the training,each rat was required to remain on the apparatus continuously ata given speed for 2 min before the speed was increased byanother increment. If a rat fell before 2 min elapsed, it was givena brief rest period (approximately 60 s) before being placedback on the apparatus. Behavioral testing commenced onlywhen all rats were able to walk on the rotorod apparatus at aspeed of 32 rpm for 2 min for two consecutive training sessions.Once this criterion was met, behavioral training was suspendedfor the remainder of the study.

2.4. Behavioral testing

All test sessions were conducted using a cumulative dosingprocedure. In this procedure, a rat was removed from its homecage and injected i.p. with vehicle, and then immediatelyreturned to its cage. After a 15-min pretreatment interval, the ratwas placed on the rotorod apparatus and the latency to fall wasrecorded at both low (8 rpm) and high (32 rpm) speeds. The

Table 1Baseline (i.e., vehicle-control) latencies a(S.E.M.) to fall from the rotorodapparatus before (prechronic), during (chronic) and after (postchronic) GHBtreatment at low (8 rpm) and high (32 rpm) speeds

Condition/vehicle 8 rpm 32 rpm

PrechronicGHBb 60.00 (0.00) 46.60 (5.85)Flunitrazepamc 60.00 (0.00) 34.86 (7.60)Baclofend 60.00 (0.00) 60.00 (0.00)

ChronicGHBb 60.00 (0.00) 41.19 (6.47)Flunitrazepamc 60.00 (0.00) 39.89 (7.75)Baclofend 60.00 (0.00) 60.00 (0.00)

PostchronicGHB b 60.00 (0.00) 46.60 (5.95)Flunitrazepamc 60.00 (0.00) 32.30 (7.67)Baclofend 60.00 (0.00) 55.11 (4.22)a Maximum latency: 60 s.b GHB vehicle: distilled water.c Flunitrazepam vehicle: saline, ethanol and Alkamuls EL-620 (v/v/v=60/20/20).d Baclofen vehicle: physiological saline (0.9% NaCl).

85M.A. Smith et al. / European Journal of Pharmacology 552 (2006) 83–89

order of testing was counterbalanced across rats and at least 60 sseparated testing at each speed. If the rat failed to fall within60 s, it was removed from the apparatus and assigned a maximallatency of 60 s. Immediately following testing at both speeds,the rat was administered the lowest does of the test drug and theprocedure was repeated. At the beginning of each subsequentcomponent, the rat was administered a higher dose of the testdrug, such that the dose increased the cumulative amount ofdrug received in that session by either 0.25 or 0.5 log unit. Forinstance, the administration of 300, 700 and 700 mg/kg over thecourse of a session yielded cumulative doses of 300, 1000 and1700 mg/kg. All sessions consisted of four or five components,depending on the drug tested.

2.5. Schedule of testing

Behavioral tests were conducted twice per week in the earlyafternoon (approximately 13:00), with at least 2 days separatingeach test. Prior to chronic treatment with GHB, all rats weretested with GHB, flunitrazepam and baclofen, in that order.Following the completion of these tests, all rats were treatedchronically with 1000 mg/kg GHB, b.i.d., a dose thatapproximated its ED50 and ED90 values at the low and highspeeds, respectively. During the chronic phase of the study,GHB was administered as a single i.p. dose at 09:00 and 19:00each day. No behavioral tests were conducted during the firstweek of chronic treatment. After 1 week, behavioral testingresumed and all rats were tested with GHB, flunitrazepam,baclofen, and then retested with GHB. Rats were notadministered the maintenance dose of GHB on mornings inwhich test sessions were scheduled but received their eveningmaintenance dose at the regularly scheduled time. Followingcompletion of these tests, all rats underwent a 2-week washoutperiod during which no drugs were administered and behavioraltests were suspended. After 2 weeks, all rats were tested withGHB, flunitrazepam and baclofen following the same protocolas that used prior to chronic treatment.

2.6. Drugs

Sodium γ-hydroxybutyrate (GHB), flunitrazepam and R(+)-baclofen HCl were obtained from Sigma Chemical Co. (St.Louis, MO, USA). Doses of GHB and baclofen refer to the saltand were dissolved in distilled water and physiological saline,respectively. Doses of flunitrazepam refer to the base and weredissolved in a solution of saline, ethanol and Alkamuls EL-620(v/v/v=60/20/20). Alkamuls EL-620 (ethoxylated castor oil)was obtained from Rhone-Poulenc (Cranbury, NJ, USA). Alldrug injections were administered i.p. in a volume of 1.0–2.0 ml/kg. Vehicle-control injections were always administeredi.p. in a volume of 1.0 ml/kg.

2.7. Data analysis

Data from all test sessions are depicted as a percentage ofvehicle-control values obtained during the first component ofthat session. For each dose-effect curve, ED50 values (95%

confidence limits) were computed mathematically (least squaresmethod) by log-linear interpolation using data from eachindividual rat (Procedure 8, Tallarida and Murray, 1987).From these data, relative potency estimates (95% confidencelimits) were computed via parallel line assay (Procedure 11,Tallarida and Murray, 1987). Dose and condition (prechronicvs. chronic vs. postchronic) effects were determined viarepeated-measures analysis of variance (ANOVA), with bothdose and condition serving as within-subject variables. Thealpha level was set at 0.05 for all statistical tests.

3. Results

3.1. Baseline motor performance

Baseline (i.e., vehicle-control) latencies to fall from the rotorodapparatus differed across speeds and conditions (Table 1). All ratsreached the maximal latency of 60 s at the low speed underprechronic, chronic and postchronic conditions. Latencies weremarkedly shorter at the high speed, and performance varied as afunction of condition. At this speed, baseline latencies decreasedfrom prechronic to chronic conditions, but rebounded underpostchronic conditions to values greater than that seen underprechronic conditions. Consistent with these observations, arepeated-measures ANOVA revealed significant main effects ofspeed (F[1,6]=015.998, P=0.007) and condition (F[2,12]=5.858, P=0.017), and a significant speed×condition interaction(F[2,12]=5.858, P=0.017). No significant vehicle effects wereobserved under any condition.

3.2. GHB

Under all conditions examined, GHB produced dose-depen-dent impairments inmotor performance asmeasured by decreasesin the latency of the subjects to fall from the rotorod apparatus(Fig. 1). Analysis of ED50 values revealed that GHB was 1.1 to1.4 fold more potent at decreasing motor performance at the high

Fig. 1. Effects of cumulative doses of GHB, flunitrazepam and baclofen on motor performance before (prechronic), during (chronic) and after (postchronic)GHB treatment. Upper panels depict data collected at the low (8 rpm) speed; lower panels depict data collected at the high (32 rpm) speed. Ordinates reflect latency tofall from the rotorod apparatus as expressed as a percentage of vehicle-control values. Abscissas reflect dose of drug in g/kg (GHB) or mg/kg (flunitrazepam, baclofen).Vertical lines on data points represent the S.E.M.

86 M.A. Smith et al. / European Journal of Pharmacology 552 (2006) 83–89

speed than at the low speed (Table 2). Chronic treatment with1000 mg/kg GHB, b.i.d., failed to alter the potency of GHB todecrease motor performance when tested after 7 and 19 days ofchronic treatment. The potency of GHB was also unaltered whenexamined during a final test conducted 14 days after terminationof the chronic regimen. At both speeds, a repeated-measuresANOVA revealed a main effect of dose (low: F[2,12]=51.502,Pb0.001; high: F[2,12]=288.096, Pb0.001), but no main effectof condition or dose x condition interaction was observed.

3.3. Flunitrazepam

Flunitrazepam dose-dependently decreased motor perfor-mance at both speeds, but was 4.6 to 15.0 fold more potent atthe high speed (Fig. 1 and Table 2). The potency offlunitrazepam was not altered by chronic GHB treatment ateither speed. A repeated-measures ANOVA revealed a maineffect of dose at both speeds (low: F[3,18]=5.011, P=0.011;high: F[3,18]=13.454, Pb0.001), but no main effect ofcondition or dose x condition interaction was observed.

3.4. Baclofen

Similar to that seen with GHB and flunitrazepam, baclofendose-dependently decreased motor performance at both speeds,but was 1.4 to 1.7 fold more potent at the high speed (Fig. 1and Table 2). Chronic treatment with GHB decreased thepotency of baclofen to impair motor performance and shiftedits dose-effect curves to the right at both speeds. Analysis ofrelative potency estimates revealed that GHB treatment shiftedthe baclofen dose-effect curves 2.0 and 1.8 fold to the right atthe low and high speeds, respectively. Following termination of

the chronic regimen, the baclofen dose-effect curves shiftedback to the left and approximated those obtained underprechronic conditions. At both speeds, a repeated-measuresANOVA revealed main effects of dose (low: F[2,12]=27.910,Pb0.001; high: F[2,12]=66.983, Pb0.001) and condition(low: F[2,12] = 4.031, P= 0.046; high: F[2,12] = 5.032,P=0.026). No significant dose×condition interaction wasobserved at either speed.

4. Discussion

In the present study, the motor-impairing effects of GHB,flunitrazepam and baclofen were examined on a rotorodapparatus rotating at low (8 rpm) and high (32 rpm) speeds.Previous studies have shown that the potency and effectivenessof centrally acting drugs in this procedure depend on thedifficulty of the task, which can be manipulated by altering thespeed of the rotorod apparatus (Watzman et al., 1967; Smith andStoops, 2001; Smith et al., 2004). In the present study, dataobtained under vehicle-control conditions suggest that the twospeeds differed in their relative level of difficulty: at the lowspeed, latencies reached the maximum of 60 s under allconditions examined, whereas at the high speed, latencies weremarkedly shorter and varied significantly across prechronic,chronic and postchronic conditions.

All three drugs produced dose-dependent decreases in thelatency to fall from the rotorod apparatus, but were consistentlymore potent at the high speed than at the low speed. Thesefindings are consistent with previous studies showing thatGABA-receptor agonists are more potent and/or more effectiveat high speeds than at low speeds in this procedure (Smith andStoops, 2001; Smith et al., 2004). In the present study, the

Table 2ED50 values

aand relative potency estimates b (95% confidence limits) of test drugs before (prechronic), during (chronic) and after (postchronic) GHB treatment whentested at low (8 rpm) and high (32 rpm) speeds

Test drug/condition ED50 value (8 rpm) ED50 value (32 rpm) Relative potency (8 rpm) Relative potency (32 rpm)

GHBPrechronic 848.77 (641.62–1122.79) 642.65 (534.46–772.75) – –Chronic (7 days) 720.69 (539.36–962.97) 628.03 (561.25–702.76) 0.85 (0.56–1.24) 0.97 (0.80–1.18)Chronic (19 days) 884.08 (677.43–1153.77) 626.85 (550.23–714.14) 1.05 (0.73–1.53) 0.97 (0.80–1.20)Postchronic 716.32 (578.24–887.37) 691.96 (546.91–875.49) 0.84 (0.60–1.16) 1.09 (0.83–1.43)

FlunitrazepamPrechronic 1.29 (0.32–5.11) 0.28 (0.12–0.69) – –Chronic 5.19 (0.04–690.14) 0.27 (0.08–0.92) 1.33 (0.06–100.74) 1.13 (0.29–4.74)Postchronic 3.89 (0.46–33.02) 0.26 (0.08–0.90) 2.70 (0.55–44.43) 1.15 (0.28–4.76)

BaclofenPrechronic 6.43 (5.03–8.23) 4.51 (3.45–5.93) – –Chronic 10.75 (7.45–15.51) 7.64 (5.92–9.87) 1.97 (1.31–3.30) 1.79 (1.27–2.71)Postchronic 6.78 (4.65–9.88) 5.04 (3.37–7.53) 1.09 (0.73–1.63) 1.08 (0.71–1.64)a Values expressed in mg/kg.b Values reflect potency estimates relative to prechronic conditions as determined via parallel line assay.

87M.A. Smith et al. / European Journal of Pharmacology 552 (2006) 83–89

degree to which speed was a determinant of a drug's potencydiffered markedly across test drugs. When averaged acrossconditions, GHB was 1.2 fold more potent at the high speed,baclofen was 3.0 fold more potent at the high speed, andflunitrazepam was 12.9 fold more potent at the high speed. It isnot known why the potency of flunitrazepam was more affectedby the difficulty of the task relative to the other drugs, but itshould be noted that the dose-effect curves of flunitrazepamwere markedly shallower than those of the other drugs, whichcontributed to greater variability in its ED50 estimates (seeFig. 1 and Table 2).

Chronic administration of 1000 mg/kg GHB, b.i.d., failed toproduce tolerance to its effects on motor performance. The doseselected for chronic administration was behaviorally active, andapproximated its ED50 and ED90 values at the low and highspeeds, respectively. Moreover, GHB was given for a period oftime that is normally sufficient to produce tolerance to theeffects of centrally acting drugs (i.e., approximately 3 weeks).Although early studies failed to observe tolerance to the effectsof GHB after chronic administration, recent studies havereported tolerance to its locomotor, cataplectic and intoxicatingeffects (Bania et al., 2003; Itzhak and Ali, 2002). Notably, thesestudies report that the dose and frequency of administrationnecessary to produce tolerance are greater than those requiredfor other centrally acting drugs, with one study reportingmaximal degrees of tolerance at a dose of 2000 mg/kgadministered every 3 h (Bania et al.). It should be noted that aprevious study reported tolerance to the effects of GHB onrotorod performance in rats treated once daily with 1000 mg/kgGHB (Colombo et al., 1995). In that study, rats were treated for9 consecutive days with GHB, and tests were conducted every24 h, 60 min after administration of GHB. Importantly, ratswere given only one exposure to the rotorod before thecommencement of testing, and thus tolerance was likely theresult of learning to perform the task while impaired. It isunlikely that similar learning processes took place in the present

study, as all rats were trained to maximal levels of performanceat both speeds prior to the commencement of testing.

GHB is a chemical precursor to the amino acid neurotrans-mitter GABA, and some of its effects may be mediated, in part,by activity at GABA receptors (Maitre, 1997; Wong et al.,2004). Evidence for a role of GABAA receptors in the effects ofGHB is mixed. Although GHB does not bind to GABAA

receptors, it may function as an indirect agonist via its metabolicconversion to GABA (Hechler et al., 1997). Supporting thispossibility, the indirect GABAA agonist, diazepam, partiallysubstitutes for GHB in drug discrimination studies (Colombo etal., 1998; Carter et al., 2003). On the other hand, GHB does notreliably substitute for diazepam in diazepam-trained animals(Carter et al., 2004), suggesting that activity at GABAA

receptors plays only a limited role in its effects.In the present study, chronic administration of GHB failed to

produce cross-tolerance to the indirect GABAA agonist,flunitrazepam. Given that drugs sharing a similar mechanismof action typically produce cross-tolerance to one another, aparsimonious explanation for these findings is that GHB andflunitrazepam do not share a common mechanism of action.This explanation is tempered by the finding that GHB did notproduce tolerance to its own effects. Tolerance is presumed tobe due, in part, to compensatory responses of individual neuronsto chronic administration of an agonist (Julien, 2005). If thecompensatory responses induced by chronic GHB administra-tion were insufficient to produce tolerance to its own effects,then it would not be expected that tolerance would be conferredto another drug that acts by a similar mechanism.

In contrast to the present findings, and supporting thepossibility that GHB shares a mechanism of action withflunitrazepam, are a limited number of studies reporting cross-tolerance between GHB and GABAA receptor agonists. Forexample, in rats treated chronically with the indirect GABAA

agonist, barbital, tolerance was conferred to the sedating effectsof GHB (Gray and Taberner, 1985). Similarly, in a study

88 M.A. Smith et al. / European Journal of Pharmacology 552 (2006) 83–89

examining the development of cross-tolerance between GHBand ethanol, which is believed to have activity at GABAA

receptors, cross-tolerance developed to the motor-impairingeffects of both drugs (Colombo et al., 1995). It is not clear,however, whether these examples reflect functional adaptationsin GABAA receptors, changes in drug metabolism, or associa-tive learning processes taking place during chronic treatment(Nicholson and Balster, 2001). Further studies using higherdoses and more frequent dosing regimens of GHB will beneeded to resolve this issue.

Evidence for a role of GABAB receptors in the effects ofGHB is less equivocal. GHB is believed to function as anindirect agonist at GABAB receptors through its metabolicconversion to GABA (Hechler et al., 1997; Maitre, 1997; Wonget al., 2004). There is also evidence that GHB binds directly toGABAB receptors, where it may function as a weak agonist(Mathivet et al., 1997). In drug discrimination studies, baclofenfully substitutes for GHB in GHB-trained rats (Colombo et al.,1998; Lobina et al., 1999), and GHB substitutes partially forbaclofen in baclofen-trained rats (Carter et al., 2004). Moreover,the GABAB receptor antagonist, CGP 35348, is equally potentat antagonizing the discriminative stimulus effects of GHB andbaclofen (Carter et al., 2003), suggesting that the effects of GHBare mediated, in large part, by activation of GABAB receptors.

Consistent with the hypothesis that GHB functions as aGABAB receptor agonist, tolerance was conferred to the motor-impairing effects of baclofen at both speeds, as revealed byrightward shifts in its dose-effect curves during GHB treatment.To our knowledge, this is the first demonstration that chronicadministration of GHB produces cross-tolerance to the effects ofa GABAB agonist. It is important to note that the baclofen dose-effect curves shifted back to the left following the discontinu-ation of treatment, indicating that the changes in potency wereindeed a consequence of chronic GHB administration.

Although previous studies have not examined cross-tolerancebetween GHB and GABAB receptor agonists, a small number ofstudies have examined the development of cross-tolerancebetween the chemical precursors of GHB and GABAB agonists.In an early study, tolerance was conferred to the effects ofbaclofen in mice treated with the GHB precursor, γ-butyrolac-tone (GBL); however, tolerance was not conferred to the effectsof GBL in baclofen-treated mice (Gianutsos and Moore, 1978).In a more recent study, tolerance was conferred to the effects ofbaclofen in rats treated chronically with 1,4-butanediol underconditions in which tolerance failed to develop to 1,4-butanediolitself (Eckermann et al., 2004). Such findings, coupled with thepresent data, suggest that chronic exposure to GHB may lead tofunctional adaptations in GABAB receptors, resulting in thedevelopment of cross-tolerance to GABAB agonists.

In summary, chronic administration of GHB produced cross-tolerance to the GABAB receptor agonist, baclofen, underconditions in which tolerance was not conferred to the indirectGABAA receptor agonist, flunitrazepam. These data areconsistent with the in vivo behavioral profile of GHB, whichreveals a greater role for GABAB receptors than for GABAA

receptors in its behavioral effects. Further studies will be neededto determine whether cross-tolerance between GHB and

baclofen is symmetrical, and whether higher maintenancedoses or more frequent dosing intervals is sufficient for GHBto produce cross-tolerance to agonists at GABAA receptors.

Acknowledgements

This study was supported by Davidson College and USPublic Service Grant DA14255 from the National Institute onDrug Abuse. The authors thank Amy Becton for expert animalcare and maintenance.

References

Bania, T.C., Ashar, T., Press, G., Carey, P.M., 2003. Gamma-hydroxybutyricacid tolerance and withdrawal in a rat model. Acad. Emerg. Med. 10,697–704.

Byrnes, J.J., Miller, L.G., Greenblatt, D.J., Shader, R.I., 1993. Chronicbenzodiazepine administration. XII. Anticonvulsant cross-tolerance butdistinct neurochemical effects of alprazolam and lorazepam. Psychophar-macology 111, 91–95.

Carter, L.P., Flores, L.R., Wu, H., Chen, W., Unzeitig, A.W., Coop, A., France,C.P., 2003. The role of GABAB receptors in the discriminative stimuluseffects of gamma-hydroxybutyrate in rats: time course and antagonismstudies. J. Pharmacol. Exp. Ther. 305, 668–674.

Carter, L.P., Unzeitig, A.W.,Wu, H., Chen,W., Coop, A., Koek,W., France, C.P.,2004. The discriminative stimulus effects of gamma-hydroxybutyrate andrelated compounds in rats discriminating baclofen or diazepam: the role ofGABA(B) and GABA(A) receptors. J. Pharmacol. Exp. Ther. 309, 540–547.

Chodera, A., Cenajek, D., Chodera, R., Nowakowska, E., Szczawinska, K.,1984. Development of tolerance to benzodiazepines. II. Baclofen effect onthe activity of rats during tolerance to the sedative action of nitrazepam andoxazepam. Acta Physiol. Pol. 35, 301–309.

Colombo,G.,Agabio, R., Lobina, C., Reali, R., Fadda, F., Gessa,G.L., 1995.Crosstolerance to ethanol and gamma-hydroxybutyric acid. Eur. J. Pharmacol. 273,235–238.

Colombo, G., Agabio, R., Lobina, C., Reali, R., Gessa, G.L., 1998. Involvement ofGABA(A) andGABA(B) receptors in themediation of discriminative stimuluseffects of gamma-hydroxybutyric acid. Physiol. Behav. 64, 293–302.

Cook, C.D., Aceto, M.D., Coop, A., Beardsley, P.M., 2002. Effects of theputative antagonist NCS382 on the behavioral pharmacological actions ofgammahydroxybutyrate in mice. Psychopharmacology 160, 99–106.

Della Pietra, G., Illiano, G., Capano, V., Rava, R., 1966. In vivo conversion ofgamma-hydroxybutyrate into gamma-aminobutyrate. Nature 210, 733–734.

Drug Enforcement Agency, 2000. Schedules of controlled substances: additionof gamma-hydroxybutyric acid to Schedule I. Fed. Regist. 65, 13235–13238.

Eckermann, K.A., Koek, W., France, C.P., 2004. Chronic 1,4-butanedioltreatment in rats: cross-tolerance to gamma-hydroxybutyrate and (+/−)-baclofen. Eur. J. Pharmacol. 484, 259–262.

ElSohly, M.A., Salamone, S.J., 1999. Prevalence of drugs used in cases ofalleged sexual assault. J. Anal. Toxicol. 23, 141–146.

Feigenbaum, J.J., Howard, S.G., 1996. Gamma hydroxybutyrate is not a GABAagonist. Prog. Neurobiol. 50, 1–7.

Gianutsos, G., Moore, K.E., 1978. Tolerance to the effects of baclofen andgamma-butyrolactone on locomotor activity and dopaminergic neurons inthe mouse. J. Pharmacol. Exp. Ther. 207, 859–869.

Gray, P.L., Taberner, P.V., 1985. Evidence for GABA tolerance in barbiturate-dependent and withdrawn mice. Neuropharmacology 24, 437–444.

Hechler, V., Ratomponirina, C., Maitre, M., 1997. gamma-Hydroxybutyrateconversion into GABA induces displacement of GABAB binding that isblocked by valproate and ethosuximide. J. Pharmacol. Exp. Ther. 281, 753–760.

Hefferan, M.P., Fuchigami, T., Marsala, M., 2006. Development of baclofentolerance in a rat model of chronic spasticity and rigidity. Neurosci. Lett.May 19, [Epub ahead of print].

Institute of Laboratory Animal Resources, 1996. Guide for the Care and Use ofLaboratory Animals. National Academy Press, Washington, DC.

89M.A. Smith et al. / European Journal of Pharmacology 552 (2006) 83–89

Itzhak, Y., Ali, S.F., 2002. Repeated administration of gamma-hydroxybutyricacid (GHB) to mice: assessment of the sedative and rewarding effects ofGHB. Ann. N.Y. Acad. Sci. 965, 451–460.

Julien, R.M., 2005. A Primer of Drug Action, tenth ed. Worth Publishers, NewYork, NY.

Kam, P.C., Yoong, F.F., 1998. Gamma-hydroxybutyric acid: an emergingrecreational drug. Anaesthesia 53, 1195–1198.

Le, A.D., Khanna, J.M., Kalant, H., Grossi, F., 1986. Tolerance to and cross-tolerance among ethanol, pentobarbital and chlordiazepoxide. Pharmacol.Biochem. Behav. 24, 93–98.

Levy, R.A, Proudfit, H.K., 1977. The analgesic action of baclofen [beta-(4-chlorophenyl)-gamma-aminobutyric acid]. J. Pharmacol. Exp. Ther. 202,437–445.

Lingenhoehl, K., Brom, R., Heid, J., Beck, P., Froestl, W., Kaupmann, K.,Bettler, B., Mosbacher, J., 1999. Gamma-hydroxybutyrate is a weak agonistat recombinant GABA(B) receptors. Neuropharmacology 38, 1667–1673.

Lobina, C., Agabio, R., Reali, R., Gessa, G.L, Colombo, G., 1999. Contributionof GABA(A) and GABA(B) receptors to the discriminative stimulusproduced by gamma-hydroxybutyric acid. Pharmacol. Biochem. Behav. 64,363–365.

Maitre, M., 1997. The gamma-hydroxybutyrate signalling system in brain:organization and functional implications. Prog. Neurobiol. 51, 337–361.

Mathivet, P., Bernasconi, R., De Barry, J., Marescaux, C., Bittiger, H., 1997.Binding characteristics of gamma-hydroxybutyric acid as a weak butselective GABAB receptor agonist. Eur. J. Pharmacol. 321, 67–75.

Nicholson, K.L., Balster, R.L., 2001. GHB: a new and novel drug of abuse.Drug Alcohol Depend. 63, 1–22.

Schwartz, R.H., Milteer, R., LeBeau, M.A., 2000. Drug-facilitated sexualassault (‘date rape’). South. Med. J. 93, 558–561.

Smith, M.A., Craig, C.K., French, A.M., 2004. Agonist and antagonist effects ofbenzodiazepines on motor performance: influence of intrinsic efficacy andtask difficulty. Behav. Pharmacol. 15, 215–223.

Smith, M.A., Stoops, W.W., 2001. Sensitivity to the effects of sedative-hypnotics on motor performance: influence of task difficulty and chronicphenobarbital administration. Behav. Pharmacol. 12, 125–134.

Tallarida, R.J., Murray, R.B., 1987. Manual of Pharmacologic CalculationsWithComputer Programs. Springer, New York, NY.

Tunnicliff, G., Raess, B.U., 2002. Gamma-Hydroxybutyrate (orphan medical).Curr. Opin. Investig. Drugs 3, 278–283.

Watzman, N., Barry III, H., Kinnard Jr., W.J., Buckley, J.P., 1967. Influence ofcertain parameters on the performance of mice on the rotarod. Arch. Int.Pharmacodyn. Ther. 169, 362–374.

Vayer, P., Mandel, P., Maitre, M., 1985. Conversion of gamma-hydroxybutyrateto gamma-aminobutyrate in vitro. J. Neurochem. 45, 810–814.

Wong, C.G., Gibson, K.M., Snead III, O.C., 2004. From the street to the brain:neurobiology of the recreational drug gamma-hydroxybutyric acid. TrendsPharmacol. Sci. 25, 29–34.

Xie, X.H., Tietz, E.I., 1992. Reduction in potency of selective gamma-aminobutyric acidA agonists and diazepam in CA1 region of in vitrohippocampal slices from chronic flurazepam-treated rats. J. Pharmacol. Exp.Ther. 262, 204–211.