Abstract
Radioligand binding studies with [3H](2E)-(5-hydroxy-5,7,8,9-tetrahydro-6H-benzo[a][7]annulen-6-ylidene) ethanoic acid ([3H]NCS-382), an antagonist of γ-hydroxybutyric acid (GHB) receptor, revealed specific binding sites in the rat cerebral cortex and hippocampus. However, there was very little binding in the rat cerebellum, heart, kidney, liver, and lung membranes. Binding was rapid and reached equilibrium in about 5 min. Scatchard analysis of saturation isotherms revealed two different populations of binding sites in the rat cerebral cortex (Kd1, 795 nM,Bmax1, 25.4 pmol/mg of protein;Kd2, 21 μM;Bmax2, 178 pmol/mg of protein) as well as in the rat hippocampus (Kd1, 441 nM;Bmax1, 16.2 pmol/mg of protein;Kd2, 9.8 μM;Bmax2, 255 pmol/mg of protein). (±)Baclofen (500 μM) and γ-aminobutyric acid (100 μM) inhibited the binding only partially, whereas (+)bicuculline, muscimol, picrotoxinin, and phaclofen did not modify the binding. Interestingly, potassium chloride (100–300 mM) inhibited [3H]NCS-382 binding (34–56%), and this inhibitory effect was not affected by picrotoxinin. GHB and NCS-382 completely inhibited the [3H]NCS-382 (16 nM) binding in the rat cerebrocortical and hippocampal membranes, and NCS-382 was found to be about 10 times more potent than GHB in this regard. A variety of ligands for other receptors did not modify the [3H]NCS-382 binding, thereby suggesting selectivity of this radioligand for the GHB receptor sites in the brain. Based on these observations, [3H]NCS-382 seems to be a better radioligand than [3H]GHB for investigating the role of the GHB receptors in various pharmacological actions.
γ-Hydroxybutyric acid (GHB) is a naturally occurring substance derived from GABA and is found in micromolar concentrations in the mammalian brain as well as in several peripheral tissues (Nelson et al., 1981; Vayer and Maitre, 1988; Snead, 1996). GHB is reported to induce absence seizures (Godschalk et al., 1977; Bernasconi et al., 1992). It is also reported to cause sedation with loss of righting reflex and anesthesia (Godbout and Pivik, 1982). GHB is rapidly absorbed and crosses the blood-brain barrier easily. Thus it has been used as anesthetic adjuvant as well as for the treatment of sleep disorders and alcohol/opioid dependence (Gallimberti et al., 1994; Addolorato et al., 2000; Gessa et al., 2000). GHB also induces euphoria and leads to physical dependence following its chronic use (Addolorato et al., 2000). Besides being used as a therapeutic agent, GHB has been used nonclinically as a so called “club drug” (also referred to as a date-rape drug in the streets), and its abuse potential is on the rise. GHB is reported to cause significant changes in the spontaneous firing of dopaminergic neurons in the substantia nigra and an increase in striatal tyrosine hydroxylase activity (Morgenroth et al., 1976). It is also reported to elevate cGMP levels with an increase in inositol phosphate turnover without affecting cAMP levels in the hippocampus (Vayer and Maitre, 1989). Although GHB possesses diverse neuropharmacological effects, the exact basic mechanism of its actions is unknown. High-affinity binding sites for GHB have been demonstrated and are distributed mainly in the cerebral cortex, hippocampus, olfactory tracts, striatum, and thalamus, whereas hypothalamus, cerebellum, and pons medulla are devoid of the GHB binding sites (Benavides et al., 1982a; Snead and Liu, 1984; Snead and Nichols, 1987; Hechler et al., 1987; Castelli et al., 2000). The identification of specific neuronal synthesis (Rumigny et al., 1981a,b), specific transport (Benavides et al., 1982b), Ca2+-dependent release (Vayer and Maitre, 1988), Na+-dependent uptake mechanisms (Hechler et al., 1985), binding (Benavides et al., 1982a; Snead and Liu, 1984; Hechler et al., 1987), and cellular responses (Vayer and Maitre, 1989) for GHB in brain indicate that GHB acts as a neuromodulator or neurotransmitter. Although binding studies with [3H]GHB have revealed the presence of high- and low-affinity specific binding sites (Benavides et al., 1982a; Maitre et al., 1990), there is a report indicating that there may be a conversion of GHB into GABA, which in turn may activate the GABAB receptor (Hechler et al., 1997). There is also evidence that GHB directly activates the GABAB receptors (Bernasconi et al., 1992; Xie and Smart, 1992; Mathivet et al., 1997; Lingenhoehl et al., 1999). Recently, it has been reported that GHB at a high dose increases the GABA contents, whereas at a low dose it decreases the GABA content in the rat frontal cortex but not in the hippocampus (Gobaille et al., 1999).
(2E)-(5-Hydroxy-5,7,8,9-tetrahydro-6H-benzo[a][7]annulen-6-ylidene) ethanoic acid (NCS-382), a synthetic structural analog of GHB (Fig.1), is reported to be an antagonist of the GHB receptor sites (Maitre et al., 1990). It blocks the increase in the release of dopamine in striatum following the GHB administration in vivo. NCS-382 also elicits anticonvulsant activity against GHB and several other animal models of seizure (Maitre et al., 1990). NCS-382 has been also shown to antagonize the electrophysiological (Godbout et al., 1995) and behavioral effects of GHB (Schmidt et al., 1991; Colombo et al., 1995). Furthermore, NCS-382 displaces [3H]GHB from both the high- and low-affinity binding sites (Maitre et al., 1990; Castelli et al., 2000). At present, [3H]GHB is the only radioligand available to investigate the pharmacology of the GHB receptor. There is a need to develop another radioligand for the GHB binding sites because GHB has a direct agonistic activity for GABAB receptor (Bernasconi et al., 1992; Xie and Smart, 1992; Mathivet et al., 1997;Lingenhoehl et al., 1999), and it is also reported to interact with the GABAA receptors gated chloride channels (Snead and Nichols, 1987). Furthermore, GHB is susceptible to Na+-dependent uptake mechanisms (Hechler et al., 1985) and conversion to GABA (Hechler et al., 1997). Hence, we evaluated custom-radiolabeled NCS-382 ([3H]NCS-382) as a radioligand for the GHB binding sites in the present study.
Experimental Procedures
Radiochemical.
NCS-382 (Fig. 1) was synthesized as described earlier (Maitre et al., 1990). The condensation between 1-benzosuberone and glyoxylic acid yielded the E-ketone and a minorZ-isomer component. Crystallization with acetic acid removed the Z-isomer to give the desired E-isomer in 46% yield. The availability of the desired E-ketone allowed reduction with 3H2 to [3H]NCS-382. American Radiolabeled Chemicals (St. Louis, MO) carried this reduction with potassium hydrobromide. The tritiated product was purified by high-pressure liquid chromatography and gave a single peak with 99% purity (specific activity, 20 Ci/mmol).
Drugs.
(+)Bicuculline, muscimol, GABA, picrotoxinin, pentobarbital, 5-HT, ketanserin, NMDA, l-glutamic acid, kainic acid, acetylcholine, adenosine, clonidine, anddl-propranolol were purchased from Sigma (St. Louis, MO). (±)Baclofen and phaclofen were purchased from Sigma/RBI (Natick, MA). Flunitrazepam and Ro 15-1788 were a gift from Hoffmann-La Roche (Nutley, NJ). MK-801 was purchased from Tocris Cookson (St. Louis, MO), whereas naloxone was procured from Endo Laboratories Inc. (Chadds Ford, PA). All other chemicals and reagents were purchased from commercial suppliers.
Animals.
Adult male Sprague-Dawley rats (Harlan Bioproducts for Science, Inc., Indianapolis, IN) weighing 250 to 300 g were used. The animals were maintained at a constant room temperature (22°C) on a 12:12 h light/dark cycle. Food and water were available ad libitum. All experiments were conducted in accordance with the Declaration of Helsinki and/or with the Guide for the Care and Use of Laboratory Animals as adopted and promulgated by the U.S. National Institutes of Health.
Membrane Preparation.
Membranes were prepared as described previously (Mehta and Ticku, 1998, 2001). Briefly, the rats were decapitated, and the rat brain regions (cerebral cortex, hippocampus, and cerebellum) and peripheral tissues (heart, kidney, liver, and lung) were dissected. Tissues were stored at −80°C until use. The tissue was thawed and homogenized in ice-cold 0.32 M sucrose, pH 7.4 (20 ml/g of tissue), and centrifuged at 1000g for 10 min at 4°C. The supernatant was then centrifuged at 140,000g for 30 min at 4°C to obtain the mitochondrial plus microsomal (P2 + P3) fraction. This fraction was dispersed in ice-cold double-distilled deionized water and homogenized by a Brinkman Polytron (Brinkmann Instruments, Westbury, NY) at a setting of six for two 10-s bursts 10 s apart. The suspension was centrifuged at 140,000g for 30 min at 4°C. The pellet was then resuspended in ice-cold Tris buffer (50 mM, pH 7.4) and centrifuged at 140,000g for 30 min at 4°C. This step was repeated two more times. After the final centrifugation step, the pellet was suspended in a small volume of ice-cold Tris buffer (50 mM, pH 7.4) and stored frozen at −80°C overnight. On the day of assay, the tissue was thawed and washed two more times with buffer as before (140,000g, 30 min, 4°C) and then resuspended in the buffer for use in assay.
[3H]NCS-382 Binding Assay.
In initial experiments, [3H]NCS-382 binding was measured using filtration (Mehta and Ticku, 1998, 2001) and centrifugation (Ticku et al., 1978) methods. Briefly, aliquots (0.1–0.2 mg of protein for filtration assay or 0.3–0.4 mg of protein for centrifugation assay) of membrane preparation in Tris buffer (50 mM, pH 7.4) were incubated with [3H]NCS-382 (16 nM) in triplicate at 4°C for 20 min in 1 ml of total volume. Nonspecific binding was determined using NCS-382 (1 mM). In the filtration assay, the binding reaction was terminated by separating the membrane material from the incubation medium using vacuum filtration (Brandel MB-48L membrane harvester; Brandel Inc., Gaithersburg, MD), and the samples were washed twice with 2 ml of ice-cold Tris buffer (50 mM, pH 7.4). In the centrifugation assay, the binding reaction was stopped by centrifugation (50,000g, 10 min, 4°C). The supernatant liquid was decanted away, and the vials were rapidly rinsed twice with 4 ml of ice-cold Tris buffer (50 mM, pH 7.4) without disturbing the pelleted tissue. Pellets were solubilized with 0.3 ml of Soluene-350 (Packard Instrument Co., Meriden, CT) for 4 to 6 h. Scintillation liquid (3 ml) was added to the solubilized material (centrifugal assay) or to the filter paper (filtration assay) in the bio-vials. Radioactivity was quantified by liquid scintillation spectrometry. Initial experiments revealed that the specific binding was much higher using centrifugation assay versus filtration assay (Table1), therefore we chose centrifugation assay for subsequent experiments.
For determination of IC50 values, [3H]NCS-382 (16 nM) binding was carried out in the absence and presence of a displacer, whereas saturation isotherms (Kd andBmax determination) were constructed by addition of concentrations of unlabeled NCS-382 ranging from 10 nM to 50 μM to the incubation mixture containing [3H]NCS-382 (16 nM).
Data Analysis.
The data are expressed as mean ± S.E.M. Scatchard analysis of saturation isotherms was carried out using computerized nonlinear regression analysis method (GraphPad Prism program; GraphPad Software, San Diego, CA). IC50data were analyzed using DeltaGraph (DeltaPoint, Monterey, CA). The data were analyzed for each individual experiment, and the mean ± S.E.M. was then calculated. The statistical analysis was performed using Student's t test.
Results
Characteristics of [3H]NCS-382 Binding.
Preliminary experiments revealed that specific binding of [3H]NCS-382 (16 nM) is more pronounced when determined by centrifugal assay versus filtration assay (546 ± 11 versus 323 ± 18 fmol/mg of protein, p < 0.005, Table 1). Binding was linear with increasing protein concentrations of the tissue membranes (0.1–0.5 mg per assay tube; Fig.2). It reached equilibrium in about 5 min and remained constant for at least 1 h (Table 2). Hence, we used 0.3 to 0.4 mg of protein concentration of the tissue membrane per assay tube for a 20-min incubation time, and terminated the binding reaction by centrifugation method for subsequent experiments. Addition of chloride ion (KCl, 100–300 mM) decreased the binding of [3H]NCS-382 to the cortical membranes (Table3). However, picrotoxinin (100 μM) neither affected the [3H]NCS-382 binding nor reversed the effect of KCl on [3H]NCS-382 binding (Table 3). Regional analysis of the rat brain membranes revealed that the highest densities of binding sites were in hippocampus, and the lowest specific binding was found in the cerebellum (Table 4). There was very little binding in the peripheral tissues such as heart, kidney, liver, or lung (Table 4).
Scatchard analysis of saturation isotherms indicated a high-affinity site with a Kd1 of 795 ± 67 nM and a Bmax1 of 25.4 ± 1.3 pmol/mg of protein and a low-affinity site with aKd2 of 21.3 ± 5.3 μM and aBmax2 of 178 ± 28 pmol/mg of protein (n = 4) in the rat cortical membranes (Fig. 3; Table 5). Similar experiments in the rat hippocampal membranes (n = 3) revealed a high-affinity site with a Kd1 of 441 ± 81 nM and a Bmax1 of 16.2 ± 3.1 pmol/mg of protein and a low-affinity site with aKd2 of 9.8 ± 1.0 μM and aBmax2 of 255 ± 11 pmol/mg of protein (Fig. 3; Table 5).
Effect of Various Drugs on [3H]NCS-382 Binding.
GHB and NCS-382 completely inhibited the [3H]NCS-382 binding to the rat cortical and hippocampal membranes in a concentration-dependent manner (Fig.4; Table 6). NCS-382 was found to be about 10 times more potent than GHB in this regard (Table 6).
GABA (100 μM) and (±)baclofen (500 μM) partially inhibited the [3H]NCS-382 (16 nM) binding to the rat cortical membranes (Table 7). However, (+)bicuculline (100 μM) and muscimol (10 μM) did not affect the [3H]NCS-382 binding (Table 7). Furthermore, phaclofen (500 μM) neither inhibited the [3H]NCS-382 binding nor affected the inhibitory effect of (±)baclofen (500 μM) or GABA (100 μM) on [3H]NCS-382 binding to the rat cortical membranes (Table 7). Effect of several other categories of drugs such as pentobarbital (500 μM, n = 5), flunitrazepam (10 μM, n = 5), Ro 15-1788 (10 μM, n = 5), ethanol (100 mM, n = 4), 5-HT (1 mM,n = 4), ketanserin (100 μM, n = 3), NMDA (1 mM, n = 4), MK-801 (1 mM, n = 4), l-glutamic acid (1 mM, n = 4), glycine (100 μM, n = 3), kainic acid (1 mM,n = 4), naloxone (1 mM, n = 3), acetylcholine (100 μM, n = 3), adenosine (1 mM,n = 3), clonidine (1 mM, n = 3), anddl-propranolol (1 mM, n = 3) on the [3H]NCS-382 (16 nM) binding was investigated in the present study. None of these drugs had any significant effect.
Discussion
Although several diverse pharmacological effects are reported to be mediated through the GHB binding sites (Morgenroth et al., 1976;Godschalk et al., 1977; Godbout and Pivik, 1982; Vayer and Maitre, 1989; Bernasconi et al., 1992; Gallimberti et al., 1994; Addolorato et al., 2000; Gessa et al., 2000), these binding sites have not been characterized in detail due to availability of only [3H]GHB as a radioligand. The main problems in using [3H]GHB as a radioligand are its susceptibility to uptake mechanism (Hechler et al., 1985) and conversion into GABA (Hechler et al., 1997). Furthermore, it has a direct agonistic activity at the GABAB receptor (Bernasconi et al., 1992; Xie and Smart, 1992; Mathivet et al., 1997;Lingenhoehl et al., 1999), and it also interacts with the GABAA receptors gated chloride channels (Snead and Nichols, 1987). To overcome these problems, it is necessary to develop another radioligand for the GHB binding sites. In the present study, we evaluated custom radiolabeled NCS-382 ([3H]NCS-382), which is an antagonist of the GHB binding sites (Maitre et al., 1990). Although previous reports have indicated that [3H]GHB binding is pH dependent, with a dramatic decrease at neutral and basic pH (Benavides et al., 1982a; Snead and Liu, 1984), we investigated [3H]NCS-382 binding at a physiological pH 7.4 since there is a substantial binding at a physiological relevant pH. Our study revealed specific binding sites for this radioligand in the rat hippocampus and cerebral cortex. Interestingly, there was very little binding in the rat cerebellum in spite of the fact that cerebellum is rich in endogenous GHB (Snead and Morley, 1981) and its biosynthetic enzyme (Rumigny et al., 1981b). This observation suggests that GHB occurs merely as a metabolite of GABA in some tissues (e.g., cerebellum), whereas in other tissues it may have a functional role as a neuromodulator or neurotransmitter. Our results regarding the regional analysis of specific binding in the rat brain are consistent with earlier studies using [3H]GHB as a radioligand (Benavides et al., 1982a; Snead und Liu, 1984; Castelli et al., 2000). Scatchard analysis of saturation isotherms revealed two different populations of binding sites for NCS-382 in the rat cerebral cortex and hippocampus in the present study, thereby confirming earlier observations in a previous study using NCS-382 as a displacer and [3H]GHB as a radioligand (Maitre et al., 1990). Furthermore, we did not detect any significant binding in the rat peripheral tissues such as heart, kidney, liver, or lung.
In the present study, KCl inhibited the [3H]NCS-382 binding in the rat cortical membranes, which is consistent with an earlier report using [3H]GHB as a radioligand (Snead and Nichols, 1987). However, picrotoxinin neither affected the [3H]NCS-382 binding nor modified the inhibitory effect of KCl on the [3H]NCS-382 binding in our study. Thus [3H]NCS-382 offers an advantage over [3H]GHB in this regard because picrotoxinin is reported to affect the [3H]GHB binding (Snead and Nichols, 1987). Muscimol, bicuculline, and phaclofen, an antagonist of presynaptic GABABreceptor, did not modify [3H]NCS-382 binding. These observations indicate that [3H]NCS-382 does not bind to the GABAA and presynaptic GABAB receptors. The lack of modulatory effect of bicuculline and muscimol on [3H]NCS-382 binding is consistent with the reports indicating that GHB does not compete for the binding at the GABAA receptors sites (Enna and Maggi, 1979; Lloyd and Dreksler, 1979). Similar observations with bicuculline and muscimol have also been reported using [3H]GHB as a radioligand (Snead and Nichols, 1987). Furthermore, some of the pharmacological effects of GHB are reported to be opposite to those of the GABAAreceptors agonists (Godschalk et al., 1977; Osorio and Davidoff, 1979;Maitre et al., 1990; Mehta and Ticku, 1999). Interestingly, GABA and (±)baclofen, at concentrations ≥50 μM, caused significant inhibition of the [3H]NCS-382 (16 nM) binding in the rat cerebral cortex. Previous studies have also indicated the agonistic effects of GHB at the GABABreceptors (Bernasconi et al., 1992; Xie and Smart, 1992; Mathivet et al., 1997; Lingenhoehl et al., 1999) and several electrophysiological effects of GHB are thought to be mediated through the GABAB receptors (Xie and Smart, 1992). The partial inhibitory effect of GABA (100 μM) and (±)baclofen (500 μM) on [3H]NCS-382 (16 nM) binding (23–30%) in our present study is most likely not mediated through GABAB receptors because both of these drugs are reported to inhibit the [3H]baclofen binding to the GABAB receptors completely at about 10 μM with IC50 values of 22 to 84 nM (Bowery et al., 1985). In view of this, we did not investigate the effects of very large and pharmacologically irrelevant concentrations of GABA (>100 μM) and (±)baclofen (>500 μM) on [3H]NCS-382 binding. However, it is still possible that NCS-382 has some affinity for the postsynaptic GABAB receptors in addition to the GHB binding sites. Notably, the GHB binding sites and GABABreceptors are distinct because these receptors are reported to have a different anatomical distribution in the rat brain (Snead, 1996). Molecular layer of the cerebellum is reported to be rich in the GABAB receptors (Bowery et al., 1987), whereas the GHB binding sites are almost absent in this brain region (Benavides et al., 1982a). Ontogeny of the GHB binding sites are also distinctly different from that of GABAB receptors (Snead, 1994). Furthermore, baclofen has no affinity for the GHB binding sites (Snead, 1996).
GHB and NCS-382 inhibited the binding of [3H]NCS-382 to the rat cerebrocortical and hippocampal membranes completely, and NCS-382 was approximately 10 times more potent compared with GHB in this regard. However, it is not possible to compare the Kd values of [3H]NCS-382 binding in our study with the reported Kd values obtained using [3H]GHB as a radioligand in earlier studies because earlier studies were performed at physiologically irrelevant acidic pH to obtain higher binding. Although Scatchard analysis of [3H]NCS-382 binding in the present study as well as a previous report using [3H]GHB as a radioligand and NCS-382 as a displacer (Maitre et al., 1990) indicate two different populations of binding sites for NCS-382, the displacement curves in the present study indicate that NCS-382 and GHB inhibit the [3H]NCS-382 (16 nM) binding by interacting with only one population of binding sites. It is probably due to the small contribution of the high-affinity site, which constitutes only 6 to 12% of total binding sites as revealed by Scatchard analysis in our present study. A variety of ligands for other receptors did not modify [3H]NCS-382 binding, thereby suggesting selectivity of this radioligand for the GHB receptor sites in the brain. In conclusion, [3H]NCS-382 is a better radioligand than [3H]GHB for investigating the pharmacology of the GHB binding sites because picrotoxinin does not affect the [3H]NCS-382 binding, whereas the [3H]GHB binding is affected by picrotoxinin. Furthermore, NCS-382 is about 10 times more potent than GHB. Availability of [3H]NCS-382 will facilitate investigation of the role of the GHB receptors in various pharmacological actions, tolerance, and physical dependence.
Acknowledgments
We thank Professor Charles France for helpful comments and discussion, Elena Wright for valuable technical help, and Joann Faulconer for excellent secretarial help.
Footnotes
- Abbreviations:
- GHB
- γ-hydroxybutyric acid
- GABA
- γ-aminobutyric acid
- NCS-382
- (2E)-(5-hydroxy-5,7,8,9-tetrahydro-6H-benzo[a][7]annulen-6-ylidene) ethanoic acid
- 5-HT
- 5-hydroxytryptamine
- NMDA
- N-methyl-d-aspartate
- MK-801
- (−)-5-methyl-10,11-dihydro-5H-dibenzo[a,d]cyclohepten-5,10-imine maleate
- Ro 15-1788
- 8-fluoro-5,6-dihydro-5-methyl-6-oxo-4H-imidazo[1,5-a][1,4]benzodiazepine-3-carboxylic acid ethyl ester
- Received July 13, 2001.
- Accepted September 4, 2001.
- The American Society for Pharmacology and Experimental Therapeutics