Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

-Hydroxybutyrate (GHB) is a neurotransmitter in brain and an emerging drug of abuse (Roth and Giarman, 1966; Doherty et al., 1978). Widespread abuse of GHB was first reported in 1990 and has increased in recent years as evidenced by a dramatic increase in the number of patients treated for GHB intoxication (Bernasconi et al., 1999; Nicholson and Balster, 2001; Mason and Kerns, 2002). Exogenously administered GHB can induce amnesia, sedation, absence seizures, coma, and death (Roth and Suhr, 1970; Snead and Liu, 1993). GHB has been used clinically as an anesthetic, to treat alcoholism, and has been recently approved by the Food and Drug Administration for the treatment of narcolepsy (Mason and Kerns, 2002).

Although GHB is widely abused, its pharmacological profile differs from other drugs of abuse. GHB does not substitute for the discriminative stimulus effects of heroin, phencyclidine (PCP), ethanol, d-amphetamine, pentobarbital, or triazolam (Beardsley et al., 1996; Woolverton et al., 1999; Metcalf et al., 2001). Conditioned place preference has been reported for GHB in rats; however, many more exposures to GHB were required to induce preference as compared with other drugs of abuse (Martellotta et al., 1997). Drug-naïve mice self-administer GHB (Martellotta et al., 1998), whereas rhesus monkeys trained to self-administer either methohexital or PCP do not (Beardsley et al., 1996; Woolverton et al., 1999).

It is likely that multiple mechanisms contribute to the effects of GHB because it binds to multiple sites in brain and because it is metabolized to and from -aminobutyric acid (GABA) in vivo (Roth, 1970; Doherty et al., 1975; Gold and Roth, 1977; Snead et al., 1989). GHB binds to low- and high-affinity sites in cerebral cortex and hippocampus (Snead and Liu, 1984; Mehta et al., 2001), and it has been suggested that the low-affinity sites are GABA_B receptors. High-affinity [³H]GHB binding does not correlate with GABA_B receptors with regard to their ontogeny and regional distribution (Snead, 1996).

One approach for studying drugs with atypical mechanisms of action is drug discrimination, largely because of its pharmacological selectivity (Shannon and Holtzman, 1976; Mansbach and Balster, 1991; Terry et al., 1994). Among the small number of studies on GHB with these procedures, a range of conditions (gender, strain, reinforcer, task, dose, and route) have been used to study the discriminative stimulus effects of GHB. For example, female CFN-strain rats responded for water (Winter, 1981), whereas male Long-Evans rats traversed a T-maze for food (Colombo et al., 1995, 1998); smaller doses (i.p.) were studied in the former, larger doses (i.g.) in the latter. Previous studies in rats (Winter, 1981; Colombo et al., 1995, 1998; Metcalf et al., 2001) and pigeons (C.P.F., unpublished observation) discriminating GHB have implicated GABA receptors in the discriminative effects of GHB. In rats discriminating 200 mg/kg GHB i.p. (Winter, 1981) the GABAergic compounds baclofen, muscimol, and chlordiazepoxide occasioned GHB-appropriate responding whereas the non-GABAergic compounds PCP, lysergic acid diethylamide, d-amphetamine, and apomorphine did not. In rats discriminating 300 or 700 mg/kg GHB i.g. (Colombo et al., 1998) the GABAergic compounds baclofen and diazepam occasioned GHB-appropriate responding whereas dizocilpine and WIN 55,212-2 did not. The first goal of this study was to confirm and extend results from other laboratories, with a more widely used two-lever, fixed-ratio (FR) schedule of responding for food. Evaluation of compounds under these conditions should facilitate interpretation of the results within the context of a drug discrimination literature that is largely founded on this procedure. Second, although GABA is clearly important in the behavioral actions of GHB, several important issues are unresolved, including the relative importance of GHB receptors in the GHB discriminative stimulus. Thus, a second goal of this study was to examine the role of GABA and other receptors in the GHB discriminative stimulus by examining receptor-selective agonists and antagonists in combination. This is particularly important for GHB receptors, as antagonism of GHB is not unanimously obtained with NCS-382. A third goal was to characterize two metabolic precursors of GHB, 1,4-butanediol (1,4-BDL) and -butyrolactone (GBL); 1,4-BDL is reported to be used by GHB abusers, although the relative interchangeability of these compounds has not been studied.

Rats discriminating 200 mg/kg GHB were used to evaluate the following: metabolic precursors 1,4-BDL and GBL; GABA_A agonist muscimol; GABA_A positive modulators diazepam, pentobarbital, and pregnanolone; GABA_B agonist baclofen; µ-opioid agonist morphine; N-methyl-D-aspartate antagonist ketamine; and the GHB antagonist NCS-382. Time course was determined for GHB, 1,4-BDL, and GBL. The benzodiazepine antagonist flumazenil, GABA_B antagonist CGP 35348, and GHB receptor antagonist NCS-382 were studied in combination with GHB and with other pharmacologically relevant compounds.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animals. Adult male Sprague-Dawley rats (Harlan Sprague-Dawley, Inc., Indianapolis, IN; n = 8) were housed individually in 45 × 24 × 20 cm plastic cages containing rodent bedding (Sani-chips, Harlan Teklad, Madison, WI) in a colony room maintained on a 12:12 light/dark cycle (experiments conducted in the light period). Rats were fed 5 to 16 g of chow (Rat Sterilizable Diet, Harlan Teklad) after daily experimental sessions to maintain body weights. Individual weights were initially decreased to 80% of the free-feeding weight and were allowed to increase up to 350 g according to normal growth curves established for Sprague-Dawley rats. Water was available continuously in the home cage. All rats were experimentally naïve prior to training. Animals were maintained and experiments were conducted in accordance with the Institutional Animal Care and Use Committee, The University of Texas Health Science Center at San Antonio, and with the 1996 Guide for the Care and Use of Laboratory Animals (Institute of Laboratory Animal Resources on Life Sciences, National Research Council, National Academy of Sciences).

Operant Chambers. For daily sessions animals were transported from the colony room to an adjacent room supplied with white noise. Experimental sessions were conducted in sound-attenuating, well ventilated enclosures (Model ENV-022M and ENV-008CT; MED Associates Inc., St. Albans, VT), which contained an operant chamber of which three sides were Plexiglas and the fourth side was a stainless steel response panel equipped with two metal levers 11.5 cm apart. The floor of the operant chamber was a grid comprising 19 rods that were 4.8 mm in diameter, spaced 1.6 cm apart, and oriented parallel to the response panel. A 2.5-cm diameter translucent circle that could be illuminated was located on the response panel above each lever and a 5 cm × 5 cm opening located equidistant between the two levers was available for food pellet delivery. Food pellets (45 mg; PJAI-0045; Noyes Precision Pellets, Research Diets Inc., New Brunswick, NJ) were delivered from a food hopper external to the operant chamber, but within the enclosure. Data were collected using MED-PC IV software (MED Associates, Inc.) and a PC interface.

Experimental Sessions. Rats were trained to discriminate 200 mg/kg GHB from saline using a two-lever, food-reinforced procedure under a FR schedule. Discrimination training began after the animal received 100 food pellets on either the right or the left lever under a concurrent schedule of continuous reinforcement. In subsequent sessions the response requirement was systematically increased (by 1 or 2) to 10, a pretreatment (PT) period was introduced, and the duration of daily sessions was shortened from 60 min to 30 min. Terminal training sessions consisted of a 15-min PT period, during which the operant chamber was dark and responses had no programmed consequence, followed by a 15-min response period, in which the lights above both levers were illuminated and 10 responses (FR10) on one of the two levers resulted in the delivery of a food pellet. A response on the incorrect lever reset the FR requirement. The injection that was administered immediately prior to the 15-min PT determined whether responding on the right or left lever was reinforced; injection volumes were between 0.28 and 2.5 ml/kg. The lever that resulted in food pellet delivery following an injection of GHB was counterbalanced across subjects so that responding on the right lever was reinforced for four animals and on the left lever for the other four animals following the training dose of GHB. Experimental sessions were conducted 7 days a week and the order of training sessions was generally double alternation (e.g., drug, drug, saline, saline).

Drugs. Drugs were administered i.p. (pH 6 to 8) and dissolved or diluted in sterile water or saline unless otherwise noted. Compounds studied included GHB (sodium salt), pentobarbital sodium, muscimol, (±)baclofen, GBL, and 1,4-BDL (Sigma-Aldrich Corp., St. Louis, MO), morphine sulfate (National Institute on Drug Abuse, Research Technology Branch, Rockville, MD), ketamine hydrochloride (Ketaset; Fort Dodge Laboratories Inc., Fort Dodge, IA), diazepam (Sigma/RBI, Natick, MA), pregnanolone (Steraloids, Newport, RI), flumazenil (a gift from F. Hoffmann LaRoche Ltd., Basel, Switzerland), NCS-382 (sodium salt), and CGP 35348 (sodium salt). NCS-382 and CGP 35348 were synthesized as previously described (Maitre et al., 1990; Froestl et al., 1995). Diazepam was dissolved in vehicle consisting of 70% Emulphor (EL-620, a polyoxyethylated vegetable oil; GAF Corporation, Linden, NJ), 20% sterile water, and 10% ethanol (by volume). Pregnanolone was dissolved in 45% -cyclodextrin (Sigma-Aldrich) in sterile water. Flumazenil was dissolved in vehicle consisting of 50% saline, 40% propylene glycol, and 10% ethanol (by volume). Doses are expressed as the weight of the free base, except where indicated otherwise.

Data Analysis. The percentage of responses on the GHB lever during the response period is plotted as drug-appropriate responding (%DR). Rate of lever pressing on both levers is plotted in responses per second. Data are reported as the average ± S.E.M. for animals whose response rate was greater than 20% of their control vehicle rate. When a rat responded at a rate that was less than 20% of its vehicle control rate, discrimination data from that test were not included in the average. The control rate of responding was the average of the response rates from the 5 previous saline training days. Drugs (except for the antagonists flumazenil, CGP 35348, and NCS-382) were studied up to doses that markedly decreased the rate of responding. During the course of these studies two animals died from causes unrelated to the experiment.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

GHB Discrimination. The median number of days to meet testing criteria was 35 (range: 23-41). GHB occasioned dose-dependent increases in drug-appropriate responding up to doses that significantly decreased rate of responding (Fig. 1). Following the training dose of 200 mg/kg GHB, an average of 83.1 ± 12.5% drug-appropriate responding was observed. GHB precursors 1,4-BDL and GBL occasioned dose-dependent increases in GHB lever responding up to doses that significantly decreased rates of responding (Fig. 1). Doses of 178 and 320 mg/kg 1,4-BDL occasioned 77.4 ± 14.6% and 89.7 ± 7.9% GHB-appropriate responding, respectively. The largest dose of 1,4-BDL tested (320 mg/kg) decreased rate of responding to 43.1% of the vehicle control rate. In contrast, GBL elicited a maximum average of 49.8% drug-appropriate responding, but only at a dose that decreased the rate of responding to 14.4% of the vehicle control rate (Fig. 1).

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Effects of 1,4-BDL () and GBL () in rats trained to discriminate 200 mg/kg GHB () from saline. The percentage of responses on the drug-appropriate lever (%DR; top panel) and the rate of responding (responses/sec; bottom panel) are plotted as a function of dose. Data points and error bars represent the mean ± S.E.M. for 7 or 8 animals. The points above V represent data obtained after administration of the appropriate vehicle.

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Effects of baclofen (), diazepam (), and pregnanolone () in rats discriminating 200 mg/kg GHB from saline. Discrimination data for 3.2 mg/kg diazepam, n = 4 (of 8 animals tested); 5.6 mg/kg pregnanolone, n = 4 (of 7 animals tested). See Fig. 1 for other details.

Time Course Studies. Discriminative stimulus effects of GHB were evident 15 min after administration (Fig. 3). At 30 min postinjection, the discriminative stimulus effects of 200 mg/kg GHB were still present as evidenced by 74.1 ± 18.6% drug-appropriate responding. However, 60 min after administration of 200 mg/kg GHB, drug lever responding was similar to that observed with vehicle (Fig. 3). 1,4-BDL exhibited a longer duration of action than GHB, as evidenced by substantial drug-appropriate responding 60 and 90 min after administration (99.3 ± 0.3% and 40.7 ± 17.8%, respectively; Fig. 3). The largest dose of GBL that did not decrease responding (32 mg/kg) did not occasion greater than 40% GHB-appropriate responding when longer pretreatment intervals were used (upper panel, Fig. 3).

View larger version (21K): [in this window] [in a new window]   Fig. 3.   Time course for GHB (), 1,4-BDL (), and GBL (). The percentage of responses on the drug-appropriate lever (%DR; top panel) and the absolute rate of responding (responses/sec; bottom panel) are plotted as a function of pretreatment time. Data points and error bars represent the mean ± S.E.M. for 7 animals.

Antagonism Studies. The purported GHB antagonist NCS-382 in combination with 200 mg/kg GHB decreased drug-appropriate responding from 83.1 ± 12.5% (control) to 38.0 ± 16.9% at a dose of 32 mg/kg NCS-382 (Fig. 4). Doses of 10 and 32 mg/kg NCS-382 also attenuated the modest rate decreasing effects of 200 mg/kg GHB. At 56 and 100 mg/kg NCS-382 (in combination with GHB), drug-appropriate responding and rate of responding were not markedly different from 200 mg/kg GHB alone (Fig. 4). When administered alone, NCS-382 occasioned substantial responding on the GHB lever (56 and 100 mg/kg) and dose-dependently decreased rate of responding (Fig. 4, Table 1). A maximum of 49.5 ± 19.1% drug-appropriate responding was obtained with 100 mg/kg NCS-382.

View larger version (16K): [in this window] [in a new window]   Fig. 4.   Effects of NCS-382 alone () and in combination () with 200 mg/kg GHB in rats trained to discriminate 200 mg/kg GHB from saline. Discrimination data for 32 mg/kg NCS-382; n = 6 (of 7 animals tested). 100 mg/kg NCS-382; n = 6 (of 8 animals tested). 178 mg/kg NCS-382; n = 4 (of 7 animals tested). 56 mg/kg NCS-382 + 200 mg/kg GHB; n = 5 (of 8 animals tested). See Fig. 1 for other details.

View larger version (15K): [in this window] [in a new window]   Fig. 5.   Effects of flumazenil in combination with diazepam () and GHB () in rats trained to discriminate 200 mg/kg GHB from saline. Discrimination data for 1.0 mg/kg flumazenil in combination with diazepam; n = 4 (of 8 animals tested). 3.2 flumazenil in combination with diazepam; n = 5 (of 7 animals tested). See Fig. 1 for other details.

View larger version (15K): [in this window] [in a new window]   Fig. 6.   Effects of CGP 35348 in combination with baclofen () and GHB () in rats trained to discriminate 200 mg/kg GHB from saline. Discrimination data for 10 mg/kg CGP 35348 in combination with baclofen; n = 6 (of 7 animals tested). See Fig. 1 for other details.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

With reports of increased recreational use of GHB as well as the recent marketing of GHB for the treatment of narcolepsy there has been a renewed interest in understanding the mechanism(s) by which this molecule exerts abuse-related and therapeutic effects. Since its isolation, much has been learned about GHB, although the mechanisms that mediate the behavioral actions of GHB are not fully described. Among the many procedures that have been used to study GHB is drug discrimination, a procedure that has a high degree of pharmacologic specificity. Previous discrimination studies with GHB in rats (Winter, 1981; Colombo et al., 1995, 1998; Metcalf et al., 2001) suggested a role for GABAergic mechanisms; however, the extent to which GABA or other receptors mediate these effects of GHB and the extent to which other metabolically related compounds share effects with GHB were not fully explored. The current study established stimulus control with GHB in rats responding under conditions that have been used widely in drug discrimination research (two-lever, FR, food-maintained responding); for compounds that were studied in both the current procedure and with other procedures, substitution data appear relatively consistent across different subjects, doses, routes of administration, gender, strain, and reinforcer. Moreover, the time required to establish stimulus control was similar to what has been reported for drug discrimination in rats with GHB or other drugs (Winter, 1981; Engel et al., 2001).

Abuse of other GHB-related compounds has reportedly increased, perhaps, because GHB was designated Schedule I (Mason and Kerns, 2002). This study provides clear evidence that one of these metabolic precursors of GHB, 1,4-BDL, shares discriminative stimulus effects with GHB. 1,4-BDL does not bind to high-affinity GHB sites, GABA_A, or GABA_B receptors and is believed to exert its effects by conversion to GHB (Roth and Giarman, 1968; Benavides et al., 1982; Snead and Liu, 1984; Carter et al., 2002), an hypothesis that is consistent with the similar discriminative stimulus effects of these two compounds in rats as well as reports of 1,4-BDL use by human GHB abusers (Mason and Kerns, 2002). Discriminative stimulus effects of GHB were evident 15 min after administration, whereas an equi-effective dose of 1,4-BDL produced comparable effects 30 to 90 min after administration. The comparatively longer time course of 1,4-BDL is also consistent with 1,4-BDL being converted to GHB (Roth and Giarman, 1968; Maxwell and Roth, 1972; Snead et al., 1989). The longer duration of action of 1,4-BDL could be preferable in future studies with chronic dosing.

1,4-BDL and GBL are metabolized to GHB by different enzymes (alcohol dehydrogenase and peripheral lactonases, respectively); thus, differences in pharmacokinetics and distribution of enzymes could account for the comparatively smaller effect obtained with GBL (Roth et al., 1966; Maxwell and Roth, 1972; Gold and Roth, 1977; Snead et al., 1989). The limited effects that were obtained in this study with GBL, as well as with morphine and ketamine, are consistent with effects of the same or similar (PCP) drugs in female rats discriminating the same dose of GHB while under other conditions (Winter, 1981).

NCS-382 antagonizes some effects of GHB including increases in cGMP levels and inositol phosphate turnover in hippocampus, release of dopamine in striatum, and seizures in rats and mice (Maitre et al., 1990). However, under other conditions NCS-382 not only fails to antagonize GHB it can enhance some actions of GHB (Schmidt-Mutter et al., 1998; Carai et al., 2001; Cook et al., 2002); the mechanism of this enhancement is not established. In the current study smaller doses of NCS-382 partially attenuated GHB-lever responding; however, larger doses of NCS-382 failed to antagonize the discriminative stimulus effects of GHB and when these doses were studied alone they elicited considerable responding on the GHB-associated lever. Similarly under other conditions smaller doses of NCS-382 (25, 50 mg/kg) antagonized the discriminative stimulus effects of GHB, whereas larger doses (300, 500 mg/kg) enhanced the sedative/hypnotic effects of GHB (Colombo et al., 1995; Schmidt-Mutter et al., 1998; Carai et al., 2001). These biphasic actions of NCS-382 further support the notion that the GHB discriminative stimulus comprises multiple mechanisms. The likelihood of obtaining antagonism of GHB with NCS-382 appears to depend on dose as well as behavioral endpoint, as NCS-382 might be an antagonist at some receptors (e.g., GHB) and an agonist at others (e.g., GABA).

GABA_B receptors appear to play an important role in the discriminative stimulus effects of GHB. Baclofen occasioned greater than 84% GHB-lever responding in 7 of 8 animals, whereas the positive GABA_A modulators diazepam and pregnanolone occasioned this level of responding in fewer animals (4 of 8 and 4 of 7, respectively) and only at doses that markedly decreased responding. The inability of flumazenil to antagonize the discriminative stimulus effects of GHB is consistent with GHB not binding to benzodiazepine receptors (Serra et al., 1991). On the other hand, the selective GABA_B receptor antagonist CGP 35348 attenuated GHB-lever responding occasioned by either baclofen or GHB, confirming a significant GABA_B component in these effects of GHB.

It is not clear whether the GABA_B-like activity of GHB results from it binding directly to GABA_B receptors or through the conversion of GHB to GABA. GHB and its antagonist NCS-382 do not alter [³H]GABA binding to GABA_B receptors and GABA_B receptor ligands do not affect [³H]GHB binding (Snead, 1996). GHB is less effective at displacing [³H]GABA from GABA_B receptors in the presence of the GHB dehydrogenase inhibitors valproate and ethosuximide (Hechler et al., 1997), further indicating that GHB is metabolized to GABA to act at GABA_B receptors. More recently (±)baclofen and GABA were shown to displace 20-30% of [³H]NCS-382 binding in rat cerebral cortex (Mehta et al., 2001). Further evidence that GHB can act directly at GABA_B receptors was provided by Lingenhoehl et al. (1999), who showed that GHB activates recombinant GABA_BR1/R2 receptors in Xenopus oocytes. Thus, in vivo GHB might act both directly at GABA_B receptors and indirectly by its conversion to GABA.

GHB does not appear to interact directly with GABA_A receptors. It does not displace [³H]muscimol, [³H]flunitrazepam, or [³⁵S]TBPS (tert-utylbicyclo-phosphorothionate) from the GABA_A receptor complex (Serra et al., 1991). Conversely, [³H]NCS-382 is not displaced by GABA_A receptor ligands such as muscimol and (+)bicuculline (Mehta et al., 2001). However, behavioral studies indicate that GHB exerts some effects through the GABA_A receptor complex. In previous studies diazepam, muscimol, and chlordiazepoxide occasioned substantial drug-appropriate responding (Winter, 1981; Colombo et al., 1998), although the doses required for these effects markedly decreased behavioral output (e.g., rate). Similarly, in the current study GHB-like responding occurred only with doses of diazepam that markedly decreased responding. Flumazenil reversed the rate decreasing effects of diazepam and GHB; however, it is unclear whether these effects were attributable to functional antagonism at benzodiazepine receptors or a general increase in rate occasioned by flumazenil. One hypothesized mechanism by which GHB could indirectly modulate the GABA_A receptor complex is by increasing brain concentrations of neurosteroids (Barbaccia et al., 2002). If discriminative stimulus effects of GHB are mediated by neuroactive steroids, then neuroactive steroids that positively modulate the GABA_A receptor complex should occasion GHB-appropriate responding as should other positive modulators (e.g., barbiturates). Pregnanolone (and pentobarbital) elicited little GHB-appropriate responding, further suggesting only a minor role for GABA_A systems in these effects of GHB.

This study had several goals. First, stimulus control was established with GHB under conditions that are used widely in drug discrimination research and substitution data generated with this procedure support results from previous studies that used different parameters and subjects. The generality of the GHB stimulus, therefore, appears to extend across a range of experimental conditions. Second, the role of GABAergic, especially GABA_B, systems was firmly established both by substitution of baclofen for GHB and, more importantly, by unequivocal antagonism of both baclofen and GHB by CGP 35348. Although GABA_A systems might also contribute to these effects of GHB, their contribution is comparatively small. Third, results from this study strongly suggest that the inconsistent literature on NCS-382 with GHB is due to GHB, and possibly NCS-382, exerting effects through more than one mechanism. Finally, greater difficulty in obtaining GHB is thought to have stimulated increased recreational use of pharmacologically related compounds such as 1,4-BDL. This study demonstrates a qualitative similarity between GHB and its precursor 1,4-BDL. To the extent that the pharmacodynamics of 1,4-BDL and GHB are qualitatively the same, 1,4-BDL might be expected to have multiple mechanisms of action and to have abuse liability similar to that of GHB.