Allosteric modulation of [3H]gabapentin binding by ruthenium red

Allosteric modulation of [3H]gabapentin binding by ruthenium red

Neuropharmacology 39 (2000) 1267–1273 www.elsevier.com/locate/neuropharm Allosteric modulation of [3H]gabapentin binding by ruthenium red Michele T. ...

133KB Sizes 0 Downloads 80 Views

Neuropharmacology 39 (2000) 1267–1273 www.elsevier.com/locate/neuropharm

Allosteric modulation of [3H]gabapentin binding by ruthenium red Michele T. Taylor, Douglas W. Bonhaus

*

Roche Bioscience, Neurobiology Unit, Center for Biological Research, 3401 Hillview Avenue, Palo Alto, CA 94304, USA Accepted 11 October 1999

Abstract Gabapentin is an anticonvulsant with an unknown mechanism of action. However, it has been proposed that gabapentin acts by binding to voltage-gated calcium channels. To further characterize the interaction of gabapentin with its endogenous binding site in cerebral cortex, we tested for competitive and allosteric interactions between [3H]gabapentin and a variety of calcium channel binding ligands. Most ligands for voltage- or ligand-gated calcium channels (verapamil, the ω-conotoxins MVIIC and GVIA, ryanodine, caffeine, capsaicin, MK-801) had no significant effect on [3H]gabapentin binding. However, ruthenium red, a relatively nonselective calcium channel ligand, was found to robustly modulate [3H]gabapentin binding. Ruthenium red slowed the association and dissociation kinetics of [3H]gabapentin while increasing the number of detectable binding sites. Spermine and MgCl2, which also bind to calcium channels and modulate [3H]gabapentin binding, were found to act in a similar manner. These findings support the contention that the principal endogenous binding site for gabapentin is a calcium channel; they characterize the nature of the allosteric interaction of spermine, MgCl2 and ruthenium red with this binding site; and they suggest possible mechanisms by which gabapentin may modulate calcium channel function and ultimately produce therapeutic actions.  2000 Elsevier Science Ltd. All rights reserved. Keywords: Gabapentin; Voltage-gated calcium channel; Ruthenium red; Spermine; MgCl2

1. Introduction Gabapentin (1-(aminomethyl)cylcohexane acetic acid) is a recently developed anticonvulsant that is efficacious in treating partial complex or generalized epileptic seizures (Goa and Sorkin, 1993; Bergey et al., 1997). Clinical findings further suggest that gabapentin will be useful in the treatment of neuropathic pain (Rosner et al., 1996; Rosenberg et al., 1997; Rowbotham et al., 1998; Backonja et al., 1998). However, the rational use of gabapentin for these indications, as well as the development of improved therapeutic agents based on clinical experience with gabapentin, is hindered by a lack of understanding of the drug’s mechanism of action (Taylor et al., 1998). One proposed mechanism of action for gabapentin has been that it somehow modulates the function of voltagegated calcium channels (VGCC). This idea is supported by the finding that gabapentin binds to the α2δ subunit

* Corresponding author. Fax: +1-650-851-3111. E-mail address: [email protected] (D.W. Bonhaus).

of VGCC (Gee et al., 1996; Suman-Chauhan et al., 1993; Dissanayake et al., 1997) and by the finding that the selectivity of enantiomers of 3-isobutyl-GABA for this binding site matches their relative potencies as anticonvulsant or anti-hyperalgesia agents (Field et al., 1997; Taylor et al., 1993). However, electrophysiological studies examining the functional interaction of gabapentin with VCCCs have produced inconsistent findings. Most studies have failed to detect a robust effect of gabapentin on calcium currents (Rock et al., 1993; Schumacher et al., 1998), while several others have reported modest inhibitory effects of gabapentin (Alden and Garcia, 1999; Stefani et al., 1998). These findings suggest that if gabapentin produces its therapeutic actions by modulation of calcium channels, it is doing so by interacting with only a subpopulation of calcium channels or by a mechanism not readily detectable by electrophysiological methods. As part of an effort to further characterize the endogenous gabapentin binding site, we tested for competitive and allosteric interactions between [3H]gabapentin and a variety of ligands which selectively interact with subtypes of calcium channels. These efforts failed

0028-3908/00/$ - see front matter  2000 Elsevier Science Ltd. All rights reserved. PII: S 0 0 2 8 - 3 9 0 8 ( 9 9 ) 0 0 1 9 8 - 7

1268

M.T. Taylor, D.W. Bonhaus / Neuropharmacology 39 (2000) 1267–1273

to disclose a unique interaction of gabapentin with a pharmacologically defined calcium channel subtype. However, they did reveal an allosteric interaction between gabapentin and a relatively nonselective blocker of ligand- and voltage-gated calcium channels, ruthenium red (RR) (Gomis et al., 1994; Hamilton and Lundy, 1995). Subsequent studies further demonstrated that spermine, MgCl2 and RR interact with the gabapentin binding site in a similar allosteric manner. A preliminary report of these findings has been presented elsewhere (Taylor and Bonhaus, 1998).

2. Methods 2.1. Membrane preparation Membranes were prepared as previously described by Dissanayake et al. (1997). Mouse cerebral cortex (PelFreez Biologicals, Rogers, AR) was homogenized in 10 volumes of buffer (0.32 M sucrose, 1 mM EDTA, 1 mM EGTA, 10 mM HEPES/KOH, pH 7.4 at 4°C), using a motor driven glass/Teflon homogenizer. The homogenate was centrifuged at 1000 g for 10 min. The pellet was discarded, and the supernatant centrifuged at 30,000 g for 20 min. The resulting pellet was resuspended in 10 volumes of buffer (1 mM EDTA, 1 mM EGTA, 10 mM HEPES/KOH, pH 7.4 at 4°C), stirred continuously on ice for 30 min and then centrifuged at 30,000 g for 20 min. The resulting pellet was resuspended in 3 volumes of buffer (1.25 mM EDTA, 1.25 mM EGTA, 25% glycerol, 12.5 mM HEPES/KOH, pH 7.4 at 4°C). The resuspended membranes (100 ml) were then combined with 25 ml of 2% Tween-20 solution and mixed for 1 h at 4°C. The mixture was centrifuged at 100,000 g in a swinging bucket rotor for 1 h. The soluble fraction was stored at ⫺70°C in 1-ml aliquots until used in binding experiments. 2.2. Competition binding studies [3H]gabapentin of concentration 20 nM was incubated with 15–20 µg protein in 10 mM HEPES/KOH buffer, pH 7.4, in the presence of various concentrations of competing agent. Incubations were for 30 min at 25°C. Bound radioactivity was separated from free ligand by filtration over 0.3% polyethyleneimine (PEI) pretreated GF/B filters. The filters were washed three times with ice cold 50 mM Tris–HCl (pH 7.4 at 4°C). The bound radioactivity was quantified using a Packard Microplate Scintillation Counter.

Fig. 1. Competition binding curves generated with 20 nM [3H]gabapentin: gabapentin — open triangles; S-(+)3-isobutyl-GABA — diamonds; GABA — circles; l-leucine — squares; d-leucine — closed triangles.

Table 1 Affinities of ligands for the [3H]gabapentin binding sitea Ligand

pKi

Hill slope

Gabapentin S-(+)3-isobutyl GABA l-leucine d-leucine GABA

7.5±0.03 7.3±0.06 7.2±0.01 ⬍5.0 ⬍5.0

⫺0.87±0.07 ⫺1.23±0.14 ⫺0.97±0.05 – –

a Values are the mean±SEM for three to five separate experiments. Each competition curve was generated with eight concentrations of displacing ligand, measured in duplicate.

(15–20 µg protein) for varying lengths of time at 4 or 30°C. For dissociation studies, membranes were first incubated with 20 nM [3H]gabapentin for 1 h. Dissociation was then initiated by adding unlabeled gabapentin for a final assay concentration of 10 µM. Dissociation reactions were terminated by filtration. Association and dissociation time courses were defined by curves containing at least 10 time-points. Nonspecific binding was defined with 10 µM unlabeled gabapentin. 2.4. Saturation binding studies Saturation experiments were carried out at 30°C. Curves were generated using 12 concentrations of [3H]gabapentin (0.5–160 nM) in a 4-h incubation with 60 µg protein. Nonspecific binding was defined with 10 µM gabapentin. The plates were filtered and counted as described for the competition studies.

2.3. Kinetic binding studies

2.5. Data analysis

Association studies were conducted by incubating [3H]gabapentin (20 nM) with solubilized membranes

Data from competition, kinetic and saturation binding studies were analyzed using nonlinear iterative curve fit-

M.T. Taylor, D.W. Bonhaus / Neuropharmacology 39 (2000) 1267–1273

1269

Table 2 Temperature-dependent effects of channel blocking ligands on [3H]gabapentin bindinga Stimulation of [3H]gabapentin binding at 30°C pEC50 Spermine Ruthenium red MgCl2

5.39±0.03 5.65±0.85 3.96±0.02

% Basal binding

Hill slope

804.5±43.5 610.0±13.6 359.2±13.1

2.21±0.28 5.30±2.38 13.0±0.74

Inhibition of [3H]gabapentin binding at 4°C pIC50 Spermine Ruthenium red MgCl2

5.74±0.75 5.03±0.16 4.70±0.13

% Basal binding

Hill slope

52.6±2.21 64.3±2.81 28.29±4.08

11.9±4.06 3.05±0.56 5.86±2.59

a Values are the mean±SEM for five to eight individual determinations. The % basal binding refers to the amount of binding at saturating concentrations of allosteric modulator compared to in the absence of an allosteric agent.

ting procedures (Inplot, GraphPad, San Diego, CA). IC50, K⫺1 and Kon values were obtained by direct analysis of the data. pKi values were derived from pIC50 (⫺log IC50) values (Cheng and Prusoff, 1973). Steady state Kd and Bmax values were obtained from saturation binding isotherms by fitting the total binding data to an equation incorporating parameters for Kd, Bmax and nonspecific binding (total bound=Bmax∗free ligand concentration/(Kd+free ligand concentration)+slope of nonspecific binding∗free ligand concentration) (Munson and Rodbard, 1980). Similar results were obtained if nonspecific binding was first subtracted and then the specific binding was fit to the equation for a rectangular hyperbola. Values are presented as the mean±SEM (standard error of the mean) for N independent determinations. Statistical comparisons of the effect of RR on various binding parameters were made using the unpaired Student’s t-test with a p value of ⬍0.05 being the criteria used to establish a statistically significant difference.

2.6. Materials

Fig. 2. (A) Stimulatory effects of ligands on [3H]gabapentin, measured at 30°C. (B) Inhibitory effects of ligands on [3H]gabapentin measured at 4°C. Spermine — squares; RR — triangles; MgCl2 — circles.

[3H]Gabapentin (42 Ci/mmol; ⬎95% purity) was synthesized by Mohammad Masjedizadeh at Roche Bioscience, Palo Alto, CA. Capsazepine, ω-conotoxins MVIIC and GVIA, and ω-agatoxin IVA were obtained from Alexis Biochemicals, San Diego, CA. Capsaicin, Dantrolene, Spermine, MgCl2 were obtained from Sigma– Aldrich, St Louis, MO. RR was acquired from Sigma and Fluka (Ronkonkoma, NY). Other ligands were synthesized by the Department of Medicinal Chemistry at Roche Bioscience.

1270

M.T. Taylor, D.W. Bonhaus / Neuropharmacology 39 (2000) 1267–1273

Table 3 Effects of ruthenium red on the kinetics of [3H]gabapentin bindinga Kinetic parameters Kon (min⫺1) K⫺1 (min⫺1) Steady state binding (fmol/mg)

Control 4°C 0.015±0.002 0.015±0.009 0.67±0.09

RR 4°C

Control 30°C

0.007±0.003* 0.004±0.003* 2.3±1.81

0.048±0.007 0.045+0.007 0.70±0.19

RR 30°C 0.011±0.002* 0.015±0.004* 1.25±0.26*

a Values are the mean pseudo first order association (Kon) and dissociation rate constants (K⫺1) in min⫺1. Steady state binding (fmol/mg) was obtained as the plateau value of the association curves. Values are from six separate experiments for each condition. *Indicates a statistically significant difference in the rate constant for gabapentin in the presence of RR (100 µM) compared to the respective control value (p⬍0.05 by unpaired t-test). Under these conditions, dissociation rate constants were similar to the observed association rate constants, precluding estimation of a Kd value from these kinetic parameters.

3. Results [3H]gabapentin specifically bound to solubilized membranes prepared from mouse cerebral cortex. When the membranes were incubated with 20 nM [3H]gabapentin at 25°C, specific binding, as defined with unlabeled gabapentin (10 µM) or (S+) 3-isobutyl GABA (10 µM), accounted for more than 80% of total binding. [3H]gabapentin binding was inhibited in a concentration-dependent manner by gabapentin, (S+) 3-isobutyl GABA and l-leucine. By contrast, d-leucine and GABA had low affinity (pKi⬍5.0) for this binding site (Fig. 1). The inhibition curves were adequately described by a single site competition model; hill slopes were not statistically different from unity (Table 1). This profile is similar to what has been previously reported for solubilized membranes, purified gabapentin binding proteins and, recombinant expressed α2δ subunits. Numerous calcium channel binding ligands including: caffeine; capsaicin; capsazepine; MK-801; ryanodine; verapamil; and the ω-conotoxins MVIIC and GVIA, had no marked effect (less than 20% inhibition) on [3H]gabapentin binding at concentrations of 10 µM (the neurotoxins) or 100 µM (all other compounds). The ωagatoxin, IVA inhibited approximately 30% of specific [3H]gabapentin binding but only at the highest concentration tested (10 µM). Dantrolene concentration-dependently inhibited [3H]gabapentin binding but with weak potency (pKi value of 4.6±0.2). The low potency of these interactions precluded their further evaluation. RR, spermine and MgCl2 produced temperaturedependent allosteric effects on [3H]gabapentin binding (Table 2). When the incubations were conducted at 4°C, RR, spermine and MgCl2 produced concentration-dependent effects inhibiting specific [3H]gabapentin binding. However, when the incubations were conducted at 30°C these ligands increased specific [3H]gabapentin binding (Fig. 2). The inhibition curves generated with RR, spermine and MgCl2, had hill slopes greater than unity and had plateau values that corresponded to less than 100% inhibition of specific binding. The hill slopes of the stimulation curves were also greater than unity (Fig. 2). The effects of MgCl2 and RR were not additive since,

in the presence of a saturating concentration of RR, MgCl2 produced no additional stimulatory action (data not shown). These findings are indicative of an allosteric interaction. Kinetic binding studies demonstrated that the rates of association and dissociation of [3H]gabapentin were temperature dependent and were modulated by RR. Both the association and the dissociation of the radioligand was slower at 4°C than at 30°C. RR at both 4 and 30°C further slowed both the observed association (Kon) and dissociation (K⫺1) rate constants. The plateau values for steady state [3H]gabapentin binding were increased by RR, though this only reached statistical significance in incubations conducted at 30°C (Table 3 and Fig. 3). The RR-stimulated [3H]gabapentin binding was reversible since the dissociation of pre-bound [3H]gabapentin reached a plateau level not different from nonspecific binding. Since the association (K1) and dissociation rate constants (K⫺1) were similar, Kd values could not be derived from these kinetic studies. Thus to determine whether RR was altering the affinity or number of detectable [3H]gabapentin binding sites, steady state saturation binding isotherms were generated. In the presence or absence of RR, [3H]gabapentin saturation binding isotherms were adequately described by a single site model. RR (100 µM) increased the Bmax for [3H]gabapentin binding from 3.44±0.65 to 13.3±1.9 pmol/mg protein (p⬍0.01, unpaired t-test, N=3) while having no statistically significant effect on its affinity (Kd values in the absence and presence of RR were 178.6±28.9 and 142.6±1.5 nM respectively) (Fig. 4). To determine whether RR was unmasking a pharmacologically distinct [3H]gabapentin binding site, the specificity of RR-enhanced [3H]gabapentin binding was characterized. In the presence of RR in a 30°C incubation, gabapentin (pKi 7.14±0.19), (S+) 3-isobutyl GABA (pKi 7.12±0.19) and, l-leucine (pKi 6.84±0.16) fully inhibited specific [3H]gabapentin binding with affinities similar to those found in the absence of RR. As was found in the absence of RR, d-leucine was a weak inhibitor (pKi⬍5) of RR-enhanced [3H]gabapentin binding. The hill slopes for the competition curves gen-

M.T. Taylor, D.W. Bonhaus / Neuropharmacology 39 (2000) 1267–1273

1271

Fig. 4. Saturation binding isotherm generated at 30°C in the presence (triangles) or absence (squares) of 100 µM RR.

Fig. 3. Effect of 100 µM RR (triangles) on the association of gabapentin at different temperatures. RR slowed the observed association rate constant and increased the estimate of steady state binding at each temperature examined. The rate constants and statistical analysis of these rate constants are presented in Table 3.

erated in the presence of RR were not different from unity.

4. Discussion To further characterize the endogenous gabapentin binding site in cerebral cortical membranes, a variety of calcium channel binding ligands were tested for their

ability to competitively or allosterically modulate [3H]gabapentin binding. In principal such an effort could provide insight into the specificity of gabapentin’s interaction with subtypes of voltage- or ligand-gated calcium channels. However, most ligands examined (caffeine, capsaicin, capsazepine, MK-801, ryanodine, verapamil, the ω-conotoxins MVIIC and GVIA) had no significant effect on [3H]gabapentin binding or, as in the case of dantroline (Parness and Palnitkar, 1995) and ω-agatoxin IVA (Currie and Fox, 1997), inhibited gabapentin binding only at concentrations greater than that attributable to an interaction at a specific subtype of channel. One ligand which was found to produce a robust effect on [3H]gabapentin binding was ruthenium red (RR). RR interacted with the gabapentin binding site in an allosteric manner, slowing the association kinetics of [3H]gabapentin while increasing the number of detectable sites. The mechanism of this interaction was not determined. However, one explanation could be that RR increased the proportion of binding sites in a conformation capable of binding [3H]gabapentin, while at the same time hindering the ligand’s access to and from the binding site. An alternative explanation, that RR unmasked a pharmacologically distinct binding site, seems unlikely given that RR had no effect on the affinity of [3H]gabapentin or on the specificity of competing ligands. A notable consequence of the effect of RR on [3H]gabapentin binding kinetics is that when incubations were conducted at 4°C, and steady-state conditions were not approached, the net effect of RR was to inhibit binding. Whereas, when the incubations were conducted at 30°C, and steady-state conditions essentially established, the net effect of RR was to increase [3H]gabapentin binding. These “temperature-dependent” effects were also produced by two previously described allosteric modulators of gabapentin binding, spermine and MgCl2 (Dissanayake et al., 1997; Suman-Chauhan et al., 1993) and may explain previous discordant findings of spemine either inhibiting or stimulating [3H]gabapentin binding.

1272

M.T. Taylor, D.W. Bonhaus / Neuropharmacology 39 (2000) 1267–1273

These findings further suggest that RR, spermine and MgCl2 allosterically modulate the gabapentin binding site in a similar manner. The findings that the gabapentin binding site in cerebral cortical membranes is allosterically modulated by the calcium channel ligands RR, spermine and Mg2+ (Pearson and Dolphin, 1993; Kuo and Hess, 1993; Gomez and Hellstrand, 1995; Drouin and Hermann, 1994; Sonna et al., 1996) further supports the idea that the principal endogenous binding site for gabapentin is a calcium channel subunit. However, given the variable findings from various electrophysiological studies of gabapentin’s effect on calcium channels, the biological relevance of this interaction remains unclear. There are several possible explanations for the failure of electrophysiological studies to reliably detect a robust effect of gabapentin on calcium currents. First, of course it is possible that there is no functional consequence of this interaction and that the biological actions of gabapentin have nothing to do with its binding to calcium channels. In this light there is no shortage of competing hypotheses to explain the actions of gabapentin (Taylor et al., 1998). Second, it is possible that gabapentin only interacts with a specific subpopulation of calcium channels. Recent studies have demonstrated the presence of multiple subtypes of α2δ calcium channel subunit with the distribution of [3H]gabapentin binding correlating with one of these subunits (α2δ⫺1) (Klugbauer et al., 1999). This finding could explain the apparent regional differences in the ability of gabapentin to inhibit calcium currents (Stefani et al., 1998; Alden and Garcia, 1999). Third, it is possible that gabapentin may alter the function of VGCCs in a manner not readily detectable by electrophysiological means such as by altering the longterm stability of the channel in the membrane or by altering the physical coupling of the channel to other intracellular constituents. Finally, since as shown in the current study, endogenous calcium channel ligands such as spermine and Mg2+ dramatically slow the association of gabapentin with its binding site, it is possible that the time required for the onset of action may have been greater than that which was anticipated in most previous studies. In summary these studies demonstrate that RR, like spermine and MgCl2, allosterically modulates the interaction of gabapentin with its endogenous binding site and that the mechanism by which these ligands modulate the binding site is similar. Since all three of these allosteric modulators interact with VGCCs, these findings further support the contention that the principal endogenous binding site for gabapentin is a calcium channel. These findings also demonstrate that these allosteric modulators markedly slow the interaction of gabapentin with its binding site. This may explain the previous discordant findings of both stimulatory and inhibitory actions of spermine on [3H]gabapentin binding and may

also have confounded previous studies of the functional interaction of gabapentin with its binding site.

Acknowledgements The authors thank Mohammad Masjedizadeh for the preparation of [3H]gabapentin.

References Alden, K., Garcia, J., 1999. Differential effects of gabapentin on calcium currents from neuronal and muscle cells. Biophysics Journal 75, A90. Backonja, M., Beydoun, A., Edwards, K.R., Schwartz, S.L., Fonseca, V., Hes, M., LaMoreaux, L., Garofalo, E., 1998. Gabapentin for the symptomatic treatment of painful neuropathy in patients with diabetes mellitus: a randomized controlled trial. Journal of the American Medical Association 280, 1831–1836. Bergey, G.K., Morris, H.H., Rosenfeld, W., Blume, W.T., Penovich, P.E., Morrell, M.J., Leiderman, D.B., Crockatt, J.G., LaMoreaux, L., Garofalo, E., Pierce, M., 1997. Gabapentin monotherapy: I. An 8-day, double-blind, dose-controlled, multicenter study in hospitalized patients with refractory complex partial or secondarily generalized seizures. The US Gabapentin Study Group 88/89. Neurology 49, 739–745. Cheng, Y.C., Prusoff, W.H., 1973. Relationship between inhibition constant (Ki) and the concentration of inhibitor which causes 50 percent inhibition (IC50) of an enzymatic reaction. Biochemical Pharmacology 92, 881–894. Currie, K.P., Fox, A.P., 1997. Comparison of N- and P/Q-type voltagegated calcium channel current inhibition. Journal of Neuroscience 17, 4570–4579. Dissanayake, V.U., Gee, N.S., Brown, J.P., Woodruff, G.N., 1997. Spermine modulation of specific [3H]-gabapentin binding to the detergent-solubilized porcine cerebral cortex α2δ calcium channel subunit. British Journal of Pharmacology 120, 833–840. Drouin, H., Hermann, A., 1994. Intracellular action of spermine on neuronal Ca2+ and K + currents. European Journal of Neuroscience 6, 412–419. Field, M.J., Holloman, E.F., McCleary, S., Hughes, J., Singh, L., 1997. Evaluation of gabapentin and S-(+)-3-isobutylgaba in a rat model of postoperative pain. Journal of Pharmacology and Experimental Therapeutics 282, 1242–1246. Gee, N.S., Brown, J.P., Dissanayake, V.U., Offord, J., Thurlow, R., Woodruff, G.N., 1996. The novel anticonvulsant drug, gabapentin (neurontin), binds to the α2δ subunit of a calcium channel. Journal of Biological Chemistry 271, 5768–5776. Goa, K.L., Sorkin, E.M., 1993. Gabapentin. A review of its pharmacological properties and clinical potential in epilepsy. Drugs 46, 409–427. Gomez, M., Hellstrand, P., 1995. Effects of polyamines on voltageactivated calcium channels in guinea-pig intestinal smooth muscle. Pflugers Archiv — European Journal of Physiology 430, 501–507. Gomis, A., Gutierrez, L.M., Sala, F., Viniegra, S., Reig, J.A., 1994. Ruthenium red inhibits selectively chromaffin cell calcium channels. Biochemical Pharmacology 47, 225–231. Hamilton, M.G., Lundy, P.M., 1995. Effect of ruthenium red on voltage-sensitive Ca++ channels. Journal of Pharmacology and Experimental Therapeutics 273, 940–947. Klugbauer, N., Lacinova´, L., Marais, E., Hobom, M., Hofmann, F.,

M.T. Taylor, D.W. Bonhaus / Neuropharmacology 39 (2000) 1267–1273

1999. Molecular diversity of the calcium channel α2δ subunit. Journal of Neuroscience 19, 684–691. Kuo, C.C., Hess, P., 1993. Block of the L-type Ca2+ channel pore by external and internal Mg2+ in rat phaeochromocytoma cells. Journal of Physiology 466, 683–706. Munson, P.J., Rodbard, D., 1980. LIGAND: a versatile computerized approach for characterization of ligand systems. Analytical Biochemistry 107, 220–239. Parness, J., Palnitkar, S.S., 1995. Identification of dantrolene binding sites in porcine skeletal muscle sarcoplasmic reticulum. Journal of Biological Chemistry 270, 18465–18472. Pearson, H.A., Dolphin, A.C., 1993. Inhibition of ω-conotoxin-sensitive Ca2+ channel currents by internal Mg2+ in cultured rat cerebellar granule neurones. Pflugers Archiv — European Journal of Physiology 425, 518–527. Rock, D.M., Kelly, K.M., Macdonald, R.L., 1993. Gabapentin actions on ligand- and voltage-gated responses in cultured rodent neurons. Epilepsy Research 16, 89–98. Rosenberg, J.M., Harrell, C., Ristic, H., Werner, R.A., De Rosayro, A.M., 1997. The effect of gabapentin on neuropathic pain. Clinical Journal of Pain 13, 251–255. Rosner, H., Rubin, L., Kestenbaum, A., 1996. Gabapentin adjunctive therapy in neuropathic pain states. Clinical Journal of Pain 12, 56–58. Rowbotham, M., Harden, N., Stacey, B., Bernstein, P., Magnus-Miller, L., 1998. Gabapentin for the treatment of postherpetic neuralgia: a

1273

randomized controlled trial. Journal of the American Medical Association 280, 1837–1842. Schumacher, T.B., Beck, H., Steinhauser, C., Schramm, J., Elger, C.E., 1998. Effects of phenytoin, carbamazepine, and gabapentin on calcium channels in hippocampal granule cells from patients with temporal lobe epilepsy. Epilepsia 39, 355–363. Sonna, L.A., Hirshman, C.A., Croxton, T.L., 1996. Role of calcium channel blockade in relaxation of tracheal smooth muscle by extracellular Mg2+. American Journal of Physiology 271, L251–L257. Stefani, A., Spadoni, F., Bernardi, G., 1998. Gabapentin inhibits calcium currents in isolated rat brain neurons. Neuropharmacology 37, 83–91. Suman-Chauhan, N., Webdale, L., Hill, D.R., Woodruff, G.N., 1993. Characterisation of [3H]gabapentin binding to a novel site in rat brain: homogenate binding studies. European Journal of Pharmacology 244, 293–301. Taylor, M.T., Bonhaus, D.W., 1998. Allosteric interactions between gabapentin and ruthenium red. The FASEB Journal 12, A439. Taylor, C.P., Gee, N.S., Su, T.Z., Kocsis, J.D., Welty, D.F., Brown, J.P., Dooley, D.J., Boden, P., Singh, L., 1998. A summary of mechanistic hypotheses of gabapentin pharmacology. Epilepsy Research 29, 233–249. Taylor, C.P., Vartanian, M.G., Yuen, P.W., Bigge, C., Suman-Chauhan, N., Hill, D.R., 1993. Potent and stereospecific anticonvulsant activity of 3-isobutyl GABA relates to in vitro binding at a novel site labeled by tritiated gabapentin. Epilepsy Research 14, 11–15.