Abstract
Accumulating preclinical data suggest that compounds that block the excitatory effect of glutamate on the kainate subtype of glutamate receptors may have utility for the treatment of pain, migraine, and epilepsy. In the present study, the in vitro pharmacological properties of the novel glutamate antagonist 5-carboxyl-2,4-di-benzamido-benzoic acid (NS3763) are described. In functional assays in human embryonic kidney (HEK)293 cells expressing homomeric GLUK5 or GLUK6 receptors, NS3763 is shown to display selectivity for inhibition of domoate-induced increase in intracellular calcium mediated through the GLUK5 subtype (IC50 = 1.6 μM) of kainate receptors compared with the GLUK6 subtype (IC50 > 30 μM). NS3763 inhibits the GLUK5-mediated response in a noncompetitive manner and does not inhibit [3H]α-amino-3-hydroxy-5-tertbutylisoxazole-4-propionic acid binding to GLUK5 receptors. Furthermore, NS3763 selectively inhibits l-glutamate- and domoate-evoked currents through GLUK5 receptors in HEK293 cells and does not significantly inhibit α-amino-3-hydroxy-5-methylisoxazole-4-propionic acid- or N-methyl-d-aspartate-induced currents in cultured mouse cortical neurons at 30 μM. This is the first report on a selective and noncompetitive GLUK5 antagonist.
Glutamate is the major excitatory neurotransmitter in the central nervous system and is involved in both physiological and pathological events in the brain through activation of G protein-coupled metabotropic receptors as well as a trio of ionotropic receptor families consisting of N-methyl-d-aspartate (NMDA) receptors, α-amino-3-hydroxy-5-methylisoxazole-4-propionic acid (AMPA) receptors, and kainate receptors (Collingridge and Lester, 1989; Schoepp et al., 1999). The pharmacology of NMDA and AMPA/kainate receptors has been studied thoroughly. Selective competitive NMDA antagonists such as d-2-amino-5-phosphovaleric acid (APV; Davies et al., 1981), noncompetitive antagonists such as (-)-5-methyl-10,11-dihydro-5H-dibenzo[a,d]cyclohepten-5,10-imine maleate (Wong et al., 1986), and the competitive AMPA/kainate receptor antagonists 6-cyano-2,3-dihydroxy-7-nitroquinoxaline and NBQX (Honoré et al., 1988; Sheardown et al., 1990) have aided in the understanding of the physiological role of these two groups of receptors. The pharmacological isolation of kainate receptors has been very difficult because AMPA receptor activation masks the existence of kainate receptors in essentially every central neuron, and the competitive AMPA/kainate receptor antagonists 6-cyano-2,3-dihydroxy-7-nitroquinoxaline and NBQX have provided limited information because they display very poor selectivity at AMPA versus kainate receptors (Fletcher et al., 1988; Lodge et al., 1991). However, low concentrations of NBQX have been used to isolate kainate responses in hippocampal interneurons (Mulle et al., 2000). The 2,3-benzodiazepine class of compounds such as GYKI52466 and GYKI53655 are noncompetitive antagonists showing a relatively high degree of selectivity for AMPA receptors, and these compounds have successfully been used to isolate kainate receptors (Paternain et al., 1995). Through use of these antagonists and cloned kainate receptors, it became clear that ATPA (Lauridsen et al., 1985) is a GLUK5-preferring agonist potently activating homomeric GLUK5 receptors and native dorsal root ganglion kainate receptors, but shows only weak activity at AMPA receptors, and no activity at GLUK6 homomers (Clarke et al., 1997; Wilding and Huettner, 2001).
Kainate receptors, as AMPA and NMDA receptors, are thought to be tetramers (for review, see Madden, 2002) formed by homo- or heteromeric association of the kainate receptor subunits GLUK1, GLUK2, GLUK5, GLUK6, and GLUK7.
The first compound to be described as a competitive kainate antagonist was NS102, based on its ability to block low-affinity [3H]kainate binding (Johansen et al., 1993). However, functional assays yielded contradictory results because NS102 acts at GLUK5 and GLUK6 receptors and shows selectivity in some systems (Verdoorn et al., 1994; Wilding and Huettner, 1996) but poor selectivity in others (Paternain et al., 1996). Recently, LY382884 has been reported to bind specifically to GLUK5 but not to GLUK6, GLUK7, GLUK2, or AMPA receptor subunits (Bortolotto et al., 1999). In functional tests, LY382884 inhibits kainate-evoked currents in dorsal root ganglion neurons and is approximately 100 times less potent on AMPA- and NMDA-evoked responses in hippocampal neurons (Bleakman et al., 2002).
Kainate receptors are believed to have diverse roles under both physiological and pathological conditions, and the novel pharmacological agents have enabled insights into the involvement of GLUK5 receptors in synaptic transmission and plasticity. In the processing of nociceptive information, kainate receptors are involved at several sites, including primary afferent fibers, superficial dorsal horn neurons, and intrinsic spinal horn neurons (Ruscheweyh and Sandkühler, 2002). Several studies have implicated kainate receptors (specifically, the GLUK5 subtype) in pain transmission (Procter et al., 1998; Li et al., 1999), and Simmons et al. (1998) demonstrated that the selective GLUK5 antagonist LY382884 was active in an animal model of persistent pain. More recently, GLUK5 receptors have been linked with migraine headache, and competitive GLUK5 antagonists have been reported to be active in animal models of acute migraine (Filla et al., 2002). Within the hippocampus, GLUK5-containing receptors are involved in frequency facilitation and induction of long-term potentiation, and in excitatory drives of inhibitory CA1 interneurons. In addition, GLUK5 antagonists have recently been reported to have anticonvulsant activity in animal models (Smolders et al., 2002).
Existing GLUK5 antagonists show no overt behavioral side effect at doses where the beneficial effects are observed in animal models (Simmons et al., 1998; Smolders et al., 2002).
In summary, these data suggest that the GLUK5 subtype of kainate receptors can be used as a target for the development of selective antagonists, which may provide a valuable approach for the future treatment of pain, migraine, and epilepsy.
In the present work, we report the in vitro pharmacology of a novel noncompetitive GLUK5 antagonist. In contrast to the competitive antagonist NS1209, which inhibits AMPA-induced responses in cortical neurons (Nielsen et al., 1999) and kainate-evoked responses in cells expressing GLUK5 receptors equipotently (Varming et al., 2001), NS3763 did not show significant antagonistic properties on either native AMPA or native NMDA receptors.
Materials and Methods
Materials and Drugs
[3H]Kainic acid (58 Ci/mmol) and [3H]ATPA (16 Ci/mmol) were purchased from PerkinElmer Life Sciences (Boston, MA) and Amersham Biosciences UK, Ltd. (Little Chalfont, Buckinghamshire, UK), respectively. Fluo-4-AM (cell-permeant acetoxymethyl ester of the Ca2+ indicator Fluo-4) and domoic acid was purchased from Molecular Probes Europe BV (Leiden, The Netherlands) and Tocris Cookson Inc. (Bristol, UK), respectively. GYKI52466 was purchased from Sigma/RBI (Natrick, MA). NS1209 (previously known as SPD502; Nielsen et al., 1999; Varming et al., 2001) was synthesized at NeuroSearch A/S. NS3763 was identified in a compound library purchased from Chemical Diversity Labs (San Diego, CA).
Cell culture media were obtained from Invitrogen (Roskilde, Denmark). All other chemicals were purchased from regular commercial sources and were of the purest grade available.
Cell Cultures
GLUK5- and GLUK6-Expressing Cell Lines. HEK293 cell lines stably expressing homomeric human GLUK5Q-1a and GLUK6IYQ were established as described previously (Varming et al., 2001).
The cell lines were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% (v/v) fetal calf serum, in polystyrene culture flasks (175 cm2), in a humidified atmosphere of 5% CO2, 95% air, at 37°C. Cells were cultured to 80 to 90% confluency before plating. The cells were rinsed with phosphate-buffered saline and detached from the culture flask by trypsin [0.1% (w/v)] digestion, for 5 min at 37°C. After addition of growth media, cells were resuspended by trituration and seeded at a density of 0.05 to 0.1 million cells/well in black-walled, clear-bottom 96-well plates pretreated with 0.001% (w/v) polyethyleneimine solution (75 μl/well for ≥30 min). Plated cells were allowed to proliferate for 24 h before loading with dye.
For experiments with the human GLUK5Q-2b splice variant, HEK293 cells were transiently transfected using the LipofectAMINE Plus (Invitrogen) transfection kit as described by the manufacturer. Cells were used the day after transfection.
For electrophysiological studies, cells were seeded on the day of experiment. Glass coverslips (3.5 mm), precoated with poly-l-ornithine [0.005% (w/v)] and laminine [0.002% (w/v)] were placed in Petri dishes, and 2.5 ml of cell suspension (0.1 million cells/ml) was added.
Primary Cortical Neuronal Cultures. The cultures were prepared from NMRI mice (Taconic M and B, Ry, Denmark) at day 15 to 16 of gestation as described previously (Drejer et al., 1987). Briefly, dissected tissue was chopped into 0.4-mm cubes and triturated with 0.2% (w/v) trypsin and DNase (40 μg/ml), for 10 min at 37°C. The cells were suspended (1.5 million/ml) in a slightly modified DMEM (with 23 mM glucose), which contained 10% (v/v) horse serum, 7 μM p-amino benzoate, 0.5 mM l-glutamine, 100 mU/l insulin, 0.1% (w/v) penicillin, and 19.1 mM KCl. The cell suspension was subsequently inoculated into poly-d-lysine-coated [0.01% (w/v)] 35-mm Petri dishes (2 ml/dish). Glass coverslips (3.5 mm) were placed in the dishes before coating. After 24 h in culture, the medium was replaced by medium without serum but with 1% N2 supplement. Every 3 to 4 days, the culture medium was replaced with DMEM/N2 supplement. Cells were maintained in culture for 8 to 13 days before experiments were carried out.
Fluorescence Measurements
On the day of experiment, the medium was aspirated from the wells, and 50 μl of a 2 μM Fluo-4-AM loading solution was added to each well. The plates were sealed and incubated at room temperature (20-22°C) for 60 min. The loading medium was then aspirated, and the cells were washed twice with 100 μl of buffer (10 mM HEPES, 140 mM choline chloride, 5 mM KCl, 1 mM MgCl2, and 10 mM CaCl2; pH 7.4) to remove extracellular dye. The reason for using a relatively high CaCl2 concentration was to enhance the fluorescent signal evoked secondary to activation of GLUK5 and GLUK6 receptors. Buffer (100 μl) was added to each well, and the fluorescence was measured at room temperature (excitation 488 nm, emission 510-570-nm band pass interference filter) using a fluorescent imaging plate reader (Molecular Devices, Sunnyvale, CA). Cells were preincubated for 1.5 min with test compound (50 μl) before addition of domoate (50 μl) to a final concentration of 2 μM (for GLUK5) or 0.2 μM (for GLUK6).
Stock solutions of test substances were made in ethanol or dimethyl sulfoxide, with final concentration of solvent never exceeding 0.5%.
Electrophysiological Studies
The electrophysiological measurements were performed in voltage clamps using conventional whole cell patch-clamp techniques (Hamill et al., 1981), and all data were obtained with an EPC-9 amplifier (HEKA Electronics, Lambrect, Germany) run by a Macintosh G3 computer. Experimental conditions and data acquisition were set and obtained using the PULSE-software accompanying the amplifier. Data were low pass filtered and sampled directly to the hard disk. Pipettes were pulled from borosilicate glass using a horizontal electrode puller (Zeitz Instrumente, Augsburg, Germany), and the final pipette resistance was approximately 2 MΩ when filled with internal solution and submerged in the external solution used in the experiments.
Coverslips with cultured cells were transferred to a perfusion chamber mounted on the stage of an inverted microscope supplied with Nomarski optics, and cells were continuously superfused with external solution at a rate of 2.5 ml/min.
Compounds were dissolved in external solution and applied to the patched cell through double-barreled application pipettes. The application pipettes were fabricated from theta glass tubes (1.5 mm outer diameter; WPI, Sarasota, FL). The application pipettes were mounted on a piezoelectric device (PZS-100HS; Burleigh Instruments, Quebec, Canada) connected to a piezo-driver (PZ-150M; Burleigh Instruments) driven by TTL pulses from the EPC-9 amplifier. One minute after the onset of the gravity flow, a PULSE protocol was initiated and the current was recorded three times separated by 30-s waiting periods. For transfected HEK293 cells, the duration of the recording periods was 150 ms during which the application pipette was switched to the test solution for 100 ms. For the cortical neurons, the duration of the recording period was 1.5 s, and the application pipette was moved for 1 s.
GLUK5- or GLUK6-Expressing Cells. For recordings from GLUK5- or GLUK6-expressing cells, the pipette solution contained 120 mM KCl, 31 mM KOH, 10 mM EGTA, 1.8 mM MgCl2, and 10 mM HEPES (pH 7.2). The external solution was composed of 140 mM NaCl, 5 mM KCl, 10 mM CaCl2, 1 mM MgCl2, and 10 mM HEPES (pH 7.4). After giga-seal formation (1-2GΩ) and establishment of the whole cell configuration, the cells were held at a holding potential of -60 mV. For each cell, a control response induced by 3 mM l-glutamate (GLUK5) or 0.5 mM l-glutamate (GLUK6) was recorded followed by recordings of the agonist-induced responses in the presence of increasing concentrations of NS3763. Because of reversibility of the effect of NS3763 and low, constant series resistance (<5 MΩ), several concentrations could be tested on each cell. Series resistance was compensated by 80%.
Native AMPA and NMDA Receptors in Cortical Neurons. For recordings in native AMPA and NMDA receptors in cortical neurons, the pipette solution contained 135 mM CsCl, 11 mM EGTA, 1.2 mM MgCl2, 0.5 mM CaCl2, 4 mM ATP, and 10 mM HEPES (pH 7.3). The external solution was composed of 140 mM NaCl, 2 mM KCl, 2 mM CaCl2, 25 mM d-glucose, 0.5 μM tetrodotoxin, and 10 mM HEPES (pH 7.4). After giga-seal formation (1-3 GΩ) and establishment of the whole cell configuration, the cells were held at a holding potential of -60 mV. For the measurement of AMPA responses, 4 mM MgCl2 was added to the external solution, whereas 10 μM glycine was added for the measurement of NMDA responses. Agonists (100 μM AMPA or 100 μM NMDA) were dissolved in the extracellular solution, and for each cell a control agonist response was recorded followed by recordings of the agonist-induced response in the presence of 30 μM NS3763.
For all electrophysiological measurements, the current amplitudes were measured at the peak of the responses, and the effect of NS3763 was calculated as the amplitude during compound application divided by the amplitude of the agonist-induced current evoked before the application of compound. All experiments were performed at room temperature (20-22°C).
Receptor Binding
GLUK5- and GLUK6-expressing cells were harvested and washed once with 50 mM Tris-HCl (pH 7.1) and stored at -80°C until the day of experiment. The thawed membrane pellets were resuspended in >100 volumes of ice-cold Tris-HCl buffer and centrifuged for 10 min (27,000g). The final pellets were resuspended in Tris-HCl buffer and used for binding experiments. All procedures were performed at 0 to 4°C.
Binding conditions for GLUK5 and GLUK6 were as described previously (Varming et al., 2001). Briefly, binding to GLUK5 receptors was performed using 3 nM [3H]ATPA at 46 to 84 μg of protein/assay, and GLUK6 receptors were labeled with 5 nM [3H]kainate at 22 to 27 μg of protein/assay. The samples were incubated in a final volume of 550 μl for 60 min at 2°C. Nonspecific binding was determined in the presence of 0.6 mM l-glutamate, and binding was terminated by rapid filtration. Radioactivity was determined by conventional liquid scintillation counting.
Data Analysis
In binding and functional studies, compounds were tested over a wide range of concentrations, and IC50 values and Hill coefficients were determined based on the equation Y = Bottom + (Top - Bottom)/(1 + (X/IC50)n), where Y is binding/calcium increase/current in percentage of total binding/calcium increase/current; X the concentration of test compound; and n is the Hill coefficient. The EC50 values for domoate in stimulation of intracellular calcium in HEK293 cells were determined by using the equation Y = 100· Xn/(EC50n + Xn).
Estimates of IC50 and EC50 values were calculated with the nonlinear curve-fitting program GraphPad Prism (version 2.0; GraphPad Software Inc., San Diego, CA). Ki values were calculated from IC50 values using the Cheng and Prusoff equation: Ki = IC50/(1 + (L/Kd)). The Kd values were as follows: 2.9 nM for [3H]ATPA at GLUK5 receptors and 5.7 nM for [3H]kainate at GLUK6 receptors, respectively. All results are given as means ± S.E.M.
Results
Calcium Measurements. Domoate concentration dependently increased intracellular calcium in GLUK5- and GLUK6-expressing cells with EC50 values of 1.5 ± 0.2 and 0.13 ± 0.03 μM, respectively (data not shown). Based on these potencies, the following inhibition studies were conducted at 2 μM domoate for GLUK5 and 0.2 μM for GLUK6 cells. NS3763 (Fig. 1) inhibited GLUK5-mediated responses with an IC50 value of 1.6 ± 0.2 μM, whereas no inhibition of GLUK6-mediated response was seen at concentrations up to 30 μM (n = 4; Table 1; Fig. 2). ATPA did not induce agonist responses at concentrations ranging from 0.01 to 3 μM, but the compound potently and selectively inhibited domoate-induced increase in intracellular calcium in GLUK5 cells with an IC50 value of 0.21 ± 0.03 μM. No inhibition of GLUK6-mediated responses was seen at concentrations up to 300 μM (n = 3; Table 1).
The competitive AMPA/kainate antagonist NS1209 potently inhibited domoate-induced calcium increase mediated by GLUK5 receptors with an IC50 value of 0.63 ± 0.09 μM (n = 4). GLUK6-mediated responses were inhibited with an IC50 value of 65 ± 4 μM by NS1209 (n = 4; Table 1; Fig. 2).
Electrophysiology on GLUK5 and GLUK6 Receptors. Both l-glutamate and domoate evoked concentration-dependent inward currents in cells stably expressing GLUK5 and GLUK6 receptors. The estimated EC50 values for l-glutamate were 3.6 and 1.0 mM for GLUK5 and GLUK6 (data not shown), respectively, and for the inhibition studies 3 and 0.5 mM l-glutamate were chosen as agonist concentration for GLUK5 and GLUK6, respectively. The peak inward current responses to 3 mM l-glutamate in GLUK5-expressing cells ranged from 122 to 1147 pA (n = 31). The l-glutamate-evoked currents were considerably larger in GLUK6-expressing cells ranging from 1.1 to 11.8 nA (n = 23) at 0.5 mM l-glutamate.
In the presence of 10 μM NS3763, the amplitude of the l-glutamate-evoked response in GLUK5 cells was inhibited by 44 ± 10% (n = 4), whereas no effect was seen on GLUK6-mediated responses up to 30 μM (n = 3, Fig. 3, A and B).
It was examined whether the effect of NS3763 was due to changes in the desensitization kinetics of GLUK5 receptors. For selected pairs of current responses evoked by 3 mM l-glutamate in the presence and absence of 30 μM NS3763, the decay phase of the peak-current was fitted to a double-exponential function and compared by paired t test. The decay time constant of the fast component, τfast, in the presence of NS3763 (1.75 ± 0.08 ms; n = 7) was not significantly different from the control value of 1.62 ± 0.10 ms (n = 7). Similarly, the slow component, τslow, was not different in the presence (8.87 ± 1.81 ms; n = 7) and absence (7.20 ± 0.69 ms; n = 7) of NS3763. The ratio of the two components was also unaffected by the application of the compound, the fast component accounting for 0.78 ± 0.03 (control) and 0.81 ± 0.04 (30 μM NS3763).
Domoate was a more potent agonist than l-glutamate in evoking currents in GLUK5- and GLUK6-expressing cells, and the estimated EC50 values were 7.5 and 0.4 μM, respectively (data not shown). The peak inward current responses obtained at the estimated EC50 value for domoate in GLUK5 cells ranged from 41 to 1474 pA (n = 9), and the amplitude of the response evoked by 8 μM domoate was inhibited by 63 ± 10% in the presence of 10 μM NS3763 (n = 5; Fig. 4A). NS3763 (0.01-30 μM) caused a concentration-dependent inhibition of GLUK5 responses to 8 μM domoate with an IC50 value of 1.3 μM, and a maximal inhibitory effect of approximately 60% (n = 5; Fig. 4B). Currents evoked by 30 μM domoate were inhibited 58 ± 5% (n = 6) in the presence of 10 μM NS3763.
All data reported above were obtained using the GLUK5-1a isoform, which contains 15 extra amino acids in the NH2-terminal domain and has a shorter COOH-terminal domain than the GLUK5-2b isoform reported for rat (Sommer et al., 1992) and for human by Korczak et al. (1995). To investigate whether NS3763 had different modulatory effects at GLUK5-1a and GLUK5-2b receptors, HEK293 cells transiently expressing the GLUK5-2b isoform were used. In these cells an EC50 value of 1.2 mM was obtained for l-glutamate (data not shown).
As illustrated in Fig. 5, NS3763 caused a concentration-dependent inhibition of l-glutamate responses in both isoforms; however, the drug was approximately 10-fold more potent at the GLUK5-2b (IC50 = 0.38 μM; n = 3) than the GLUK5-1a isoform (IC50 = 5.7 μM; n = 4).
In electrophysiological studies, as well as in assays measuring intracellular calcium (Fig. 2), only a partial inhibition of the responses could be obtained.
Mechanism of Action of NS3763 on GLUK5 Receptors. To investigate the mechanism of inhibition by NS3763, its effect on the concentration-response relationship for domoate was characterized. As illustrated in Fig. 6A, the maximal increase in intracellular calcium occurred at 30 μM domoate, and 3 μM NS3763 caused a reduction of the maximal response. However, the EC50 value for domoate in the presence of NS3763 (1.7 ± 0.2 μM; n = 3) was similar to the control value of 1.4 ± 0.1 μM (n = 3). In contrast, 3 μM NS1209 caused a concentration-dependent rightward shift in the concentration-response curve with no change in the maximal response (Fig. 6A); the EC50 value for domoate in the presence of 3 μM NS1209 was 7.1 ± 1.3 μM (n = 3). Thus, unlike NS1209, which interacts competitively at the domoate binding site, NS3763 inhibits domoate responses by a noncompetitive mechanism.
The noncompetitive action of NS3763 was supported by radioligand binding studies, because NS3763 did not inhibit [3H]ATPA binding to GLUK5 receptors (Table 1). In contrast, NS1209 inhibited ligand binding to GLUK5 and GLUK6 (Varming et al., 2001), whereas ATPA only displaced binding to GLUK5.
The noncompetitive mechanism of action for NS3763 was confirmed in electrophysiological studies (Fig. 6B); the EC50 value for l-glutamate in the presence of 10 μM NS3763 (3.28 mM; n = 5-9) was similar to the control value of 3.35 mM (n = 3-7; p > 0.05; paired t test).
Native AMPA and NMDA Receptors. The possible antagonistic activity of NS3763 on native AMPA- and NMDA receptors was investigated in cultured mouse cortical neurons. The peak inward current responses to 100 μM AMPA or NMDA ranged from 104 to 244 pA (n = 4) and from 525 to 1237 pA (n = 4), respectively. In the presence of 30 μM NS3763, the amplitude of the responses to 100 μM AMPA was inhibited by -0.6 ± 5.4% (n = 4; Fig. 7A). In contrast, the AMPA-induced current was blocked by 89 ± 2% in the presence of 30 μM GYKI52466, a noncompetitive AMPA antagonist (n = 4; Fig. 7A). The response to 100 μM NMDA was blocked by 10 ± 4% in the presence of 30 μM NS3763 and by 79 ± 11% by 50 μM APV (n = 4; Fig. 7B). The inhibition of AMPA- and NMDA-induced currents by NS3763 was not significant (p > 0.05; paired t test).
Discussion
The present study shows that NS3763 selectively and noncompetitively inhibits homomeric GLUK5 receptor-mediated responses. This is the first demonstration of a noncompetitive kainate antagonist.
In studies measuring intracellular calcium, NS3763 was shown to inhibit domoate-induced responses in GLUK5 receptor-expressing HEK293 cells (IC50 = 1.6 μM). ATPA and the mixed AMPA/kainate receptor antagonist NS1209 (Nielsen et al., 1999; Varming et al., 2001) also inhibited domoate-induced calcium responses in GLUK5-expressing cells with IC50 values of 0.21 and 0.63 μM, respectively. The lack of agonist response observed for ATPA at concentrations up to 3 μM and its inhibitory action on domoate-induced responses is most likely due to rapid desensitization of homomeric GLUK5 receptors (Lerma et al., 1993, 2001). Similarly, no agonist response could be obtained for l-glutamate (data not shown).
The effects of the two antagonists NS3763 and NS1209 on domoate responses, however, were different. NS1209 shifted domoate concentration-response curves to the right in a parallel manner indicative of competitive antagonism, whereas the effect of NS3763 was noncompetitive. The noncompetitive mechanism of action of NS3763 was further supported by the fact that NS3763, in contrast to ATPA and NS1209, did not inhibit [3H]ATPA binding to GLUK5 receptors.
In electrophysiological studies, NS3763 inhibited domoate-evoked currents in GLUK5-expressing cells (IC50 = 1.3 μM), with a potency similar to that determined using intracellular calcium measurements. The inhibition of agonist-evoked currents by NS3763 was, as expected from its noncompetitive mechanism of action, independent of the concentration of agonist used. Currents evoked by 30 μM domoate were inhibited by 10 μM NS3763 to the same extend (58 ± 5%) as seen at 8 μM of the agonist (Fig. 4B). Despite the different functional endpoints (current versus cytosolic calcium) in electrophysiological and imaging studies, the potency and maximal inhibitory effect (60-70%) of NS3763 against domoate-induced responses were very similar.
Currents evoked by the endogenous ligand l-glutamate were also inhibited in the low micromolar range by NS3763 with a maximal inhibitory effect varying from 30 to 50% (Figs. 5 and 6B), and the noncompetitive mechanism of action of the compound was confirmed.
Furthermore, the data indicates that NS3763 is selective for homomeric GLUK5 over homomeric GLUK6 receptors at concentrations up to 30 μM. NS3763 is somewhat less potent than NS1209 (IC50 = 0.075 μM; Varming et al., 2001) in inhibiting GLUK5 receptor-mediated currents, but whereas NS1209 displays equipotent inhibition of AMPA-induced responses in cortical neurons (Nielsen et al., 1999) and kainate evoked responses in cells expressing GLUK5 receptors (Varming et al., 2001), NS3763 did not show significant antagonistic properties on either native AMPA or native NMDA receptors at concentrations up to 30 μM. NS3763 thus shows selectivity for GLUK5 receptors and displays a selectivity profile different from NS1209. These data are the first describing a selective noncompetitive antagonist of GLUK5 receptors.
Several compounds interacting noncompetitively with AMPA receptors are known, and it is well described that AMPA receptor-mediated responses can be either diminished or enhanced by drugs acting at allosteric sites. The 2,3-benzodiazepines such as GYKI52466 and GYKI53655 are well characterized as negative allosteric modulators (Bleakman et al., 1996). The known positive allosteric modulators are benzothiadiazines (e.g., cyclothiazide) and benzoylpiperidines (e.g., aniracetam and CX516). The precise mechanism by which 2,3-benzodiazepines act as negative modulators is not known; they do not affect deactivation or desensitization and are not open channel blockers (Donevan and Rogawski, 1993; Rammes et al., 1998).
Rapid desensitization is one of the most characteristic features of kainate receptors, and modification of this feature is one of the means by which kainate receptor responses can be altered by pharmacological agents. The lectin concanavalin A has long been known to markedly reduce kainate receptor desensitization by an allosteric mechanism (Huettner, 1990), and thereby potentiates current responses of native and recombinant receptors (Huettner, 1990; Partin et al., 1993). The inhibition of l-glutamate-evoked currents by NS3763 is apparently not due to an increase in kainate receptor desensitization because the drug had no effect on the rate of l-glutamate current desensitization.
The binding site for NS3763 at homomeric GLUK5 receptors is not known. However, the compound seemed to distinguish between two isoforms of GLUK5, being 10-fold more potent in inhibiting l-glutamate-evoked currents in the GLUK5-2b than in the GLUK5-1a isoform, which has 15 additional amino acids present in the NH2 terminal (Sommer et al., 1992). The reason for these potency differences is currently being studied.
The potencies of NS3763 in inhibiting agonist-induced responses in assays measuring intracellular calcium and in electrophysiological studies were very similar, but it seems that only a partial inhibition of the responses can be obtained. The reason for this is not known, but it should be noted that the compound has limited water solubility [<0.05 mg/ml (<125 μM) at pH 7.4]. However, it cannot be excluded that NS3763 is unable to inhibit the functional responses completely due to its noncompetitive interaction with the GLUK5 receptor.
ATPA has been shown to be not completely selective for homomeric GLUK5 receptors because it also activates heteromeric GLUK5 receptors containing GLUK2 or GLUK6 subunits and heteromers consisting of GLUK6/K2 receptor subunits (Paternain et al., 2000). LY382884 has recently been reported to also be effective against heteromeric assemblies of GLUK5 and GLUK6 subunits (Smolders et al., 2002), so there is still a need for new pharmacological tools. The activity of NS3763 on heteromeric kainate receptors and the potential analgesic effect of the drug in animal models of pain are currently being studied.
Acknowledgments
We thank Anne B. Fisher, Kirsten V. Haugegaard, Kristina Christensen, and Aino Munch (Departments of Biochemical Screening and Receptor Biochemistry) for skillful technical assistance.
Footnotes
-
J.K.C. was supported by the Danish Academy of Technical Sciences.
-
DOI: 10.1124/jpet.103.062794.
-
ABBREVIATIONS: NMDA, N-methyl-d-aspartate; AMPA, α-amino-3-hydroxy-5-methylisoxazole-4-propionic acid; APV, d-2-amino-5-phosphovaleric acid; NBQX, 2,3-dihydroxy-6-nitro-7-sulfamoylbenzo[f]quinoxaline; ATPA, α-amino-3-hydroxy-5-tertbutylisoxazole-4-propionic acid; GYKI52466, 1-(4-aminophenyl)-4-methyl-7,8-methylenedioxy-5H-2,3-benzodiazepine; NS1209, 8-methyl-5-(4-(N,N-dimethylsulfamoyl)phenyl)-6,7,8,9,-tetrahydro-1H-pyrrolo[3,2-h]-isoquinoline-2,3-dione-3-O-(4-hydroxybutyric acid-2-yl)oxime; NS3763, 5-carboxyl-2,4-di-benzamidobenzoic acid; HEK, human embryonic kidney; DMEM, Dulbecco's modified Eagle's medium; GYI53655, 1-(4-aminophenyl)-3-methylcarbamyl-4-methyl-7,8-methylenedioxy-3,4-dihydro-5H-2,3-benzodiazepine; LY382884, 3S,4aR,6S,8aR-6-(4-carboxyphenyl)methyl-1,2,3,4,4a,5,6,7,8,8a-deca-hydroisoquinoline-3-carboxylic acid; NS102, 5-nitro-6,7,8,9-tetrahydrobenzo[g]indole-2,3-dione-3-oxime.
- Received November 10, 2003.
- Accepted February 24, 2004.
- The American Society for Pharmacology and Experimental Therapeutics