Journal of Neurophysiology

Effect of Conantokin G on NMDA Receptor–Mediated Spontaneous EPSCs in Cultured Cortical Neurons

Anitha B. Alex, Anthony J. Baucum, Karen S. Wilcox

Abstract

Conantokin G (Con G), derived from the venom of Conus geographus, is the most characterized natural peptide antagonist targeted to N-methyl-d-aspartate (NMDA) receptors. Although Con G is known to bind to the glutamate binding site on the NR2 subunit of the receptor, it is unclear whether it can allosterically modulate the function of the receptor through the glycine binding site on the NR1 subunit. This study was designed to evaluate the action of Con G on NMDA receptor–mediated spontaneous excitatory postsynaptic currents (sEPSCs) and its modulation by glycine in cultured cortical neurons (13–19 days in vitro) using the whole cell patch-clamp technique. Con G inhibited NMDA receptor–mediated sEPSCs in a concentration-dependent manner. Also, the potency of Con G decreased as a function of time in culture. The inhibition of EPSCs observed after application of Con G in the presence of high (10 μM) and nominal (no added) concentrations of glycine was not different at 13 days in vitro (DIV). Furthermore, similar results were obtained with experiments on Con G–induced inhibition of NMDA-evoked whole cell currents. These results indicate that glycine concentrations do not have a direct effect on Con G–induced inhibition of NMDA currents. In addition, age dependency in the action of Con G on cortical neurons in vitro suggests that this model system would be useful in examining the effects of different agonists/antagonists on native synaptic NMDA receptors.

INTRODUCTION

N-methyl-d-aspartate (NMDA) receptor (NMDAR) composition is dynamic, varying both regionally and developmentally in the brain (Lynch and Guttmann 2002; McBain and Mayer 1994; Tovar and Westbrook 1999, 2002; Wenthold et al. 2003). The NR2 subunit confers unique pharmacological properties to NMDA receptors (Brimecombe et al. 1998; Kutsuwada et al. 1992; Lynch et al. 1995; Waxman and Lynch 2005). Hence NR2 subunit specific antagonists are potentially helpful in revealing the structure and functional properties of native and recombinant NMDA receptors and may prove useful therapeutically in a variety of neurological disorders. One such class of naturally occurring pharmacological agent is Conantokins, a group of toxins produced by marine cone snails (Genus Conus). Conantokins are biochemically distinctive in their high content of the modified amino acid γ-carboxyglutamate and lack of cysteine residues (Terlau and Olivera 2004). Conantokin G (Con G), a small 17 amino acid peptide isolated from fish-hunting snail, Conus geographus, is the best-characterized conantokin (Olivera 1997), which exerts its in vivo effects through the functional inhibition of NMDA receptors (Layer et al. 2004). The potential of Con G as a neuroprotective agent in ischemic and excitotoxic brain injury, neuronal apoptosis, and Alzheimer's disease has been reported (Dave et al. 2003; Ragnarsson et al. 2002; Williams et al. 2000).

Con G is thought to be a NR2B subunit selective antagonist (Donevan and McCabe 2000; Klein et al. 2001). However, a recent report indicates that, unlike CI-1041 and other NR2B-specific antagonists, Con G blocks NMDA receptor–mediated excitatory postsynaptic currents (EPSCs) in adult hippocampal brain slices (Barton et al. 2004). The decreased potency of Con G in adult brain slices compared with the P4-P6 slices seen by Barton et al. (2004) could be attributed to the developmental increase in the expression of NR2A containing receptors. Furthermore, there is evidence to suggest that the NMDA receptor subtypes expressed at mature cortical and hippocampal synapses are predominantly NR1/NR2A and NR1/NR2A/NR2B receptors (Fu et al. 2005; Li et al. 1998; Rumbaugh and Vicini 1999; Stocca and Vicini 1998; Tovar and Westbrook 1999; Tovar et al. 2000). To further study the age-dependent decrease in the potency of Con G, we chose a model system of dissociated cortical neurons, where the action of Con G could be evaluated at immature and mature synaptic receptors and at extrasynaptic receptors as well. In this study, the effect of Con G on NMDA receptor–mediated sEPSCs in murine cortical neurons maintained in culture for 13–19 days in vitro (DIV) was assessed. Spontaneous EPSCs (sEPSCs) are the result of a complex, cyclic activation of the culture network, and the frequency, amplitude, and decay of the EPSCs are determined by variety of both network and cellular properties. Hence the effect of Con G on NMDA receptor–mediated sEPSCs was also confirmed with currents evoked by exogenous application of NMDA on cultured cortical neurons.

In addition to interacting competitively with the glutamate site, Con G may have a more complex interaction with the NMDARs (Terlau and Olivera 2004). An earlier report indicates that Con G enhances strychnine-insensitive [3H] glycine binding to rat forebrain membranes in a concentration-dependent manner (Mena et al. 1990). Moreover, there is ample evidence suggesting allosteric interactions between ligands binding at the NMDA and glycine recognition sites (Danyz and Parsons 1998; Lester et al. 1993; McBain and Mayer 1994; Priestley and Kemp 1994; Priestley et al. 1996). Studies with 5,7-DCKA, a glycine binding site antagonist, have shown that it relieves the Con G–induced block of NMDA receptors (Donevan and McCabe 2000). Furthermore, Con G, which contains a glycine at position 2, could interfere at the glycine co-agonist binding site on the NR1 subunit and induce a decrease in the affinity for glutamate (Hammerland et al. 1992). Recently, Barton et al. (2004) suggested that Con G might be acting as a partial agonist at the glycine site, resulting in a decrease in the NMDA-mediated EPSC decay times. However, whether Con G directly interacts with the glycine binding site of the NMDA receptor is not yet clear. Hence in this study, an attempt was made to evaluate whether varying glycine concentrations could affect the action of Con G on NMDA receptor–mediated sEPSCs in murine cortical neurons in vitro.

METHODS

Culture of cortical neurons

Primary cultures of cortical neurons were prepared from embryonic (E-16) mice of CF-1 strain (Charles River, Wilmington, MA), as described previously (Otto et al. 2002). Briefly, cerebral cortices were dissected out, minced, and enzymatically dissociated with 0.25% trypsin-EDTA (Sigma, St. Louis, MO). After gentle trituration, the cells were pelleted, washed, and resuspended in Dulbecco's modified Eagle's medium (DMEM) supplemented with l-glutamine, horse serum, and 100 U/ml penicillin-streptomycin (Sigma). For high-density culture, the cells were plated at 6 × 105 cells/ml on poly-l-lysine (Sigma)-coated coverslips placed in 35-mm dishes. For low-density plates, the cells were plated at 100,000 cells/ml. Cultures were maintained in a humidified incubator at 37°C and 7% CO2 for 3 wk. After 4–5 days of plating, 10 μM cytosine β-d-arabinofuranoside (Sigma) was added to the culture dishes to prevent excessive glial proliferation. The culture medium was replaced with fresh DMEM every 2–3 days. Experiments were done on neurons maintained in culture for 13–19 days. For recording sEPSCs, neurons from high-density plates were used, and for evoked NMDA currents, neurons from low-density plates were used. Animal handling and experimental procedures were in accordance with the National Institute of Health Guide for the Care and Use of Laboratory Animals.

Western blot analysis

For Western blot analysis of NR2A and NR2B NMDA receptor subunits, cortical neurons were plated at high density in 25-cm2 tissue culture flasks and maintained in culture for ≤19 DIV. 13 and 19 DIV neurons were scraped into warm PBS, pelleted, and homogenized in ice-cold 0.32 M sucrose buffer. The homogenate was centrifuged at 800g for 12 min at 4°C, and the pellet (P1) was resuspended in water and saved for later analysis. The supernatant (S1) was again centrifuged (22,000g for 15 min; 4°C), and the resulting pellet (P2) was resuspended in distilled water.

P1 and P2 homogenates were mixed with loading buffer (final concentration: 2.25% SDS, 18% glycerol, 180 mM Tris base, pH 6.8, and bromophenol blue). The samples were loaded on a 4–16% SDS-polyacrylamide gradient gel, electrophoresed, and transferred to polyvinylidene difluoride hybridization transfer membrane (NEN/Perkin Elmer, Wellesley, MA) as described previously (Baucum et al. 2004). Because of glial protein contamination, equal volumes of samples were loaded. The blots were probed with NR2A antibody first and then stripped and reprobed with NR2B, and changes in their expression were compared. Rabbit polyclonal antibodies against NR2A and NR2B subunits were obtained from Chemicon (Temecula, CA). Proteins were visualized with enhanced chemiluminescence (NEN/Perkin Elmer) using a Fluorchem SP imaging system (Alpha Innotech, San Leandro, CA).

Electrophysiological recordings and solutions

Whole cell voltage-clamp recordings were performed on murine embryonic cortical neurons maintained in culture for 13–19 DIV. Patch pipettes (2–4 MΩ) were pulled using a micropipette electrode puller (Sutter Instruments) and were filled with internal recording solution containing 153 mM CsCl2, 10 mM EGTA-CsOH, 10 mM HEPES, and 4 mM MgCl2 (290 mOsm; pH 7.3). The cells plated on glass coverslips were placed in a recording chamber perfused continuously with Mg2+-free extracellular solution containing, 142 mM NaCl, 1.5 mM KCl, 1 mM CaCl2, 20 mM sucrose, 10 mM glucose, and 10 mM HEPES (310 mOsm; pH 7.3). Except for experiments in low glycine conditions, glycine (10 μM) was added in the external solution to maximally activate the NMDA receptors. All reagents were obtained from Sigma. Working concentrations of Con G were prepared from a stock of 0.1 mM by diluting in extracellular solution.

RECORDING OF sEPSCs.

NMDA-mediated sEPSCs were recorded from 13–19 DIV cortical neurons. CNQX (10 μM) and picrotoxin (50 μM) were included in the bath solution to block non-NMDA and GABAA receptors, respectively. For experiments in low glycine conditions, NBQX (10 μM) instead of CNQX, was used, as the latter competes with glycine for the glycine binding site of the NMDA receptor (Lester et al. 1989; Yu and Miller 1995). Consecutive recordings were made in gap-free mode for 60 s with an Axopatch 200A amplifier and pClamp 8.0 software (Axon Instruments) and digitized for off-line analysis using Clampfit 8.0. Total charge transfer of EPSCs in the presence of Con G (0.01–0.3 μM) was expressed as percent of control.

RECORDING OF NMDA-GATED CURRENTS.

Recordings of evoked NMDA currents from cultured cortical neurons (13–19 DIV) were made using an EPC-7 amplifier in voltage-clamp mode. NMDA (10 μM) and glycine (10 μM) were used to evoke NMDA-gated currents. TTX (500 nM), a sodium channel blocker, was included in the external solution to block synaptic transmission. The cells were perfused with extracellular, NMDA, and NMDA + Con G solutions through a gravity-fed three-barreled microperfusion system (Isoherranen et al. 2002), positioned within 100 μm of the cell. NMDA/NMDA + Con G was applied for 0.5 s, and the currents evoked were recorded using pClamp 9.0 (Axon Instruments). Response obtained after the application of Con G was compared with the control response (NMDA alone) in each cell.

Data analysis

All data are presented as mean ± SE. P < 0.05 was considered significant. Statistical significance was determined by t-test and one-/two-way ANOVA using Graphpad Prism. For all experiments, the number of neurons recorded was at least five at each data point.

RESULTS

Age-dependent decrease in the inhibition of NMDA receptor–mediated EPSCs by Con G

To evaluate the changes in efficacy of Con G in correlation with the developmental changes in synaptic NMDA receptor subunit expression, experiments were carried out in cortical neurons cultured for 13, 16, and 19 DIV. The model of dissociated cortical neurons offers many advantages, including the ability to determine the effect of Con G on NMDA receptors at both immature and mature synapses and rapid and easy drug accessibility. NMDA receptor–mediated spontaneous excitatory postsynaptic currents (sEPSCs) were pharmacologically isolated by the addition of picrotoxin (50 μM) and CNQX (10 μM) to a Mg2+-free bath solution containing glycine (10 μM) to maximize the activation of NMDA receptors. The sEPSCs thus obtained could be abolished by the application of APV (10 μM; Fig. 1A). Conantokin G (0.01–0.3 μM) significantly decreased the amplitude and total charge transfer of NMDA receptor–mediated EPSCs in embryonic mouse cortical neurons (Fig. 1, B and C). Total charge transfer for all EPSCs in a period of 60 s was determined for each condition. Con G also induced a significant reduction in the frequency of postsynaptic currents (Fig. 1D). Con G–induced inhibition of the total charge transfer and frequency of sEPSCs were reversible.

FIG. 1.

Conantokin G (Con G) inhibits N-methyl-d-aspartate (NMDA) receptor–mediated excitatory postsynaptic currents (EPSCs) in a concentration and age-dependent manner. A: spontaneous EPSCs (sEPSCs) elicited in the presence of CNQX (10 μM) and picrotoxin (50 μM) were blocked completely by APV (10 μM). B: NMDA receptor–mediated sEPSCs recorded from cortical neurons maintained for 13 and 19 days in vitro (DIV). C: effect of Con G (0.01–0.3 μM) on total charge transfer expressed as percent of control. Efficacy of Con G decreases with increase in days neurons are maintained in vitro. D: frequency of sEPSCs at different concentrations of Con G (0.01–0.3 μM), recorded over a period of 60 s, as normalized to control. Like total charge transfer, Con G also induced a dose- and age-dependent inhibition in frequency of EPSCs. *Compared with control; #Between days (P < 0.001, two-way ANOVA, n = 8–11).

As was recently reported in hippocampal brain slices (Barton et al. 2004), the potency of Con G–induced inhibition of sEPSCs decreased as a function of the number of days the neurons were cultured (13–19 DIV; Fig. 1, B and C). At 13 DIV, Con G (0.3 μM) almost completely blocked total charge transfer of spontaneous NMDA receptor–mediated EPSCs. At 16 and 19 DIV, the Con G–induced inhibition of the total charge transfer of EPSCs was significantly lesser than the inhibition at 13 DIV. In addition to a reduction of total charge transfer, which is also indicative of frequency, Con G was found to reduce peak amplitude of the sEPSCs. At 0.3 μM of Con G, the percent of inhibition of peak amplitude at 16 days was 63.6 ± 0.8 compared with 88.4 ± 4.6 at 13 days (n = 8–11; P < 0.05, One-way ANOVA). At 19 DIV, there was only a 57.2 ± 2.7% block of EPSC amplitude with Con G (0.3 μM; n = 6–8; P < 0.01, one-way ANOVA). This decrease in amplitude of sEPSCs is correlated with both the reduction in the total charge and frequency of sEPSCs as shown in Fig. 1, C and D. The inability of Con G (0.3 μM) to completely block either frequency, amplitude, or total charge of the sEPSCs from cultures at 19 DIV confirms the decreased efficacy of Con G, which may correspond to an increased expression of NR2A containing NMDA receptors. Developmental increase in NR2A expression has been reported in cultured cortical neurons in vitro (Kew et al. 1998; Li et al. 1998; Williams et al. 1993; Zhong et al. 1994).

Con G–induced reduction of NMDA-gated currents in cortical neurons decreased with DIV

Although unlikely, the decrease in the frequency, amplitude, and total charge transfer of NMDA receptor–mediated sEPSCs seen after Con G application could be caused, in addition to its direct effects on the NMDA receptor, by unidentified effects on synaptic transmission (Layer et al. 2004). Therefore we performed experiments on NMDA currents evoked by exogenous application of NMDA (10 μM) in 13–19 DIV neurons. The neurons were perfused with external solution containing TTX (500 nM) to block synaptic transmission and glycine (10 μM) to maximize the NMDA response. Exogenous application of NMDA (10 μM) resulted in inward currents that were inhibited by Con G (0.3 μM). As was seen for sEPSCs, the efficacy of Con G inhibition of NMDA-evoked currents decreased with increased time in culture (Fig. 2). In neurons recorded at 13 DIV, Con G reduced NMDA-evoked currents to 87.4 ± 7.7% of control, whereas at 19 DIV, inhibition by Con G was only 41.4 ± 5.7% of control (n = 5 each; P < 0.05, unpaired t-test). The age dependence and the amount of inhibition of NMDA-gated currents from 13–19 DIV neurons (Fig. 2, A and B) were comparable with that seen with sEPSCs (Fig. 1, B and C).

FIG. 2.

Agonist-induced currents (10 μM NMDA/10 μM glycine) in murine cortical neurons was inhibited by Con G (0.3 μM) age dependently. A: traces of NMDA-gated currents from cortical neurons (13 and 19 DIV). Potency of Con G decreased at 19 DIV as seen with sEPSCs. B: NMDA responses in the presence of Con G (0.3 μM) from 13, 16, and 19 DIV neurons expressed as percent of control (NMDA alone). *Compared with control response (paired t-test); #between days (P < 0.05, one-way ANOVA, n = 5). Extent of inhibition of the evoked NMDA response was comparable with that seen with NMDA receptor–mediated sEPSCs in age-matched neurons. C and D: developmental changes in NMDA receptor subunit expression in cortical neurons as shown by Western blotting. Homogenates of 13 and 19 DIV neurons were loaded by volume on a 4–16% gradient gel and subjected to SDS-PAGE and immunoblotted using NR2A-specific polyclonal antibody. Hippocampal homogenate (H1) was also included as positive control. Bands were visualized by enhanced chemiluminescence using Fluorchem SP imaging system. NR2A blot was stripped and reprobed with NR2B antibody for comparison of expression. SDS-PAGE analysis has shown a major band approximately at 180 kDa for NR2A and NR2B. C: increased expression of NR2A in the P1 and P2 fractions of 19-day cell homogenate compared with the 13-day homogenate. D: expression of NR2B subunits did not change significantly between days of development.

To test the hypothesis that the decreased efficacy of Con G in older cultures was correlated to developmental changes in the NMDA receptor subunit expression, Western blot analysis of neuronal cultures were performed. Consistent with previous reports, the results from these experiments indicate that there is an increase in expression of NR2A subunits in 19 DIV neurons compared with 13 DIV (Fig. 2C) (Kew et al. 1998; Li et al. 1998; Williams et al. 1993; Zhong et al. 1994). However, no such age-dependent change in expression was seen for the NR2B subunit (Fig. 2D).

Effect of external glycine concentration on the action of Con G on NMDA receptor–mediated sEPSCs

Earlier reports indicated a greater expression of NR2A-containing receptors at mature synapses in cultured hippocampal neurons (Tovar and Westbrook 1999), and coincident with this increase, there is a decreased affinity for glycine (Kew et al. 1998; Wilcox et al. 1996). To identify whether the decreased efficacy of Con G in older cultures is caused by an interaction of Con G at the glycine binding site, recordings of sEPSCs were made from neurons in low glycine (no added) concentrations. Cortical neurons cultured for 13, 16, and 19 DIV were perfused with bath solution containing NBQX (10 μM) and picrotoxin (50 μM), and sEPSCs under these conditions were compared with those after the application of Con G (0.01–0.3 μM).

The action of Con G on NMDA receptor–mediated sEPSCs at 13 DIV was not altered by the external concentrations of glycine (Fig. 3B). However, at 16 DIV, there was a significant increase in the potency of Con G (0.01–0.3 μM) under low glycine conditions (Fig. 3C; P < 0.001, two-way ANOVA). In addition, at 19 DIV, the effect of Con G (0.01–0.3 μM) on EPSCs was statistically different, with the maximal block of NMDA-mediated currents being substantially greater in the low glycine condition (Fig. 3D; P < 0.001, two-way ANOVA). Surprisingly, the effect of Con G on 19 DIV cultures in low glycine concentration was similar to the corresponding 13 days-no glycine data (Fig. 3B). If glycine binding at the NR1 site did have modulatory effect on the action of Con G, it would be predicted that there would also be a difference in the dose–response of Con G depending on glycine concentrations on all the days tested. However, the apparent increase in the potency of Con G at 16 and 19 days in low glycine may be caused by the decreased glycine affinity of NR2A-containing receptors. Because they have a very low affinity, NR2A containing receptors may not be activated at low concentrations of glycine and that might result in a population of NR2B containing receptors that are blocked effectively by Con G. Similar observations were reported with the inhibition of NMDA currents by ifenprodil in cortical and hippocampal cultures (Kendrick et al. 1998; Kew et al. 1998).

FIG. 3.

A: NMDA receptor–mediated sEPSCs in the absence of glycine at 13 and 19 DIV. B: At 13 DIV, Con G–induced inhibition of EPSCs in the presence (10 μM) or absence of externally added (nominal) glycine was not different (P = 0.1; two-way ANOVA; n = 6–8). *Compared with control response and between concentrations (P < 0.05, one-way ANOVA). C: Con G (0.01–0.3 μM) induced a greater inhibition of NMDA receptor–mediated EPSCs from 16 DIV neurons (#P < 0.001, two-way ANOVA, n = 8) at nominal glycine concentrations. D: at 19 DIV, under low glycine conditions, there is a significant increase in the potency of Con G (#P < 0.001, two-way ANOVA; n = 5). Data at 19 days-no glycine was similar to the glycine and no-glycine data obtained from 13 DIV neurons.

Con G–induced inhibition of NMDA-gated currents is not modulated by glycine

To further explore the possible interactions of Con G at the glycine binding site, experiments were performed to assess whether varying glycine concentrations can modify the action of Con G on evoked NMDA responses in cortical neurons in vitro. Whole cell recordings from cortical neurons (13 and 19 DIV) were obtained in the presence of different concentrations of glycine (0.5–10 μM). NMDA-gated currents in the presence of Con G (0.3 μM) were compared with those evoked with NMDA alone. At 13 DIV, there was no significant difference in the action of Con G on NMDA receptors depending on the external glycine concentrations (Fig. 4, A and C). However, a significant increase in the Con G–induced inhibition of NMDA currents was seen at 19 days under low glycine conditions (Fig. 4, B and D). The increased efficacy of Con G on evoked NMDA response at 19 DIV at low glycine conditions is consistent with the data obtained with NMDA receptor–mediated sEPSCs.

FIG. 4.

Con G–induced inhibition of NMDA-gated currents in the presence of different concentrations of glycine (0.5–10 μM) A and C: effect of Con G (0.3 μM) on evoked NMDA currents in cortical neurons cultured for 13 DIV under low glycine conditions were not different from its corresponding high concentration data (n = 5 each, P > 0.05, one-way ANOVA). B and D: at 19 DIV, increase in the potency of Con G was seen at low glycine concentrations (*P < 0.05, Student's t-test, n = 5 each) similar to that observed with sEPSCs.

DISCUSSION

Recordings of NMDA receptor–mediated sEPSCs in dissociated cell culture provide an excellent and rapid developmental screen to assess the efficacy of novel compounds at synaptic NMDA receptors. This is especially true for compounds with slow on-off kinetics, where preapplication is required if effects on endogenously applied agonists are to be observed. In addition, in instances where only small quantities of compound are available, the ease of drug access for synaptic sites makes dissociated cell preparations superior to brain slices, where slow diffusion of drugs to synaptic sites requires longer equilibration times, greater amounts of solution, and therefore greater amounts of compound. This study was designed to evaluate the action of Conantokin G on NMDA receptor–mediated sEPSCs in dissociated neuronal cell culture and determine if there were interactions with this compound at the glycine-binding site of the NMDA receptor. Our results revealed that there was a dose-dependent decrease in NMDA receptor–mediated sEPSCs by Con G (0.01–0.3 μM) and that the Con G–induced inhibition decreased as a function of days that the neurons were maintained in culture. In addition, the data presented here suggest that Con G does not modulate activity of the NMDA receptor through direct interactions at the glycine binding site.

Con G, in increasing concentrations, can significantly and reversibly decrease the total charge transfer, peak amplitude, and frequency of NMDA receptor–mediated sEPSCs in neurons cultured for 13–19 DIV (Fig. 1). Previous work with Con G suggests that its actions are selective for the NMDA receptor (Layer et al. 2004). Therefore the effect of Con G on the frequency of the sEPSCs is most likely caused by a dampening of excitability of the neural circuit as a consequence of antagonizing the NMDA receptor, rather than a nonspecific effect on synaptic transmission. This is similar to what was observed with the NMDA selective antagonist, APV (Fig. 1A). This hypothesis is further supported by the results of the experiments that showed that the effect of Con G on NMDA-evoked responses, obtained in the presence of TTX, was comparable with that of the sEPSCs.

The potency of Con G decreases with time in culture, so that by day 19, the efficacy of Con G is dramatically reduced (Figs. 1 and 2). These results are consistent with the decreased potency of Con G on NMDA-mediated EPSCs observed in adult rat hippocampal brain slices compared with P4–P6 slices (Barton et al. 2004). The decreased efficacy of Con G on 19 DIV neurons is most likely caused by a developmental switch in the type of NR2 subunits that are expressed at the synapse. Faster deactivation kinetics and decreased sensitivity to ifenprodil and glycine are associated with increased NR2A expression in cortical and hippocampal neurons with time in culture (Kew et al. 1998; Ming et al. 2002; Thomas et al. 2006; Williams et al. 1993; Zhong et al. 1994). Increased expression of NR1/NR2A and NR1/NR2A/NR2B receptors at mature synapses has also been reported (Armentia and Sah 2003; Rumbaugh and Vicini 1999; Stocca and Vicini 1998; Thomas et al. 2006; Tovar and Westbrook 1999). Immunoprecipitation studies also suggest that the majority of NMDA receptors in adult rat cortex are triheteromeric NR1/NR2A/NR2B receptors (Luo et al. 1997). Our results from Western blotting studies further support the developmental increase in NR2A expression in 19 DIV neurons (Fig. 2, C and D). Therefore at 19 DIV, sEPSCs may be mediated through synaptic NMDA receptors containing an NR2A subunit. Con G is reported to be a NR2B-specific antagonist (Donevan and McCabe 2000; Klein et al. 2001), and the potency of inhibition varies with the type of NR1 splice variant present (Klein et al. 2001; Ragnarsson et al. 2006). However, the ability of Con G (0.3 μM) to block 60% of the NMDA-mediated EPSCs and NMDA-gated currents in 19 DIV neurons suggests that Con G can also bind to NR1/NR2A/NR2B containing receptors. Indeed, Con G–induced inhibition of triheteromeric (NR1/NR2A/NR2B) NMDA receptors expressed in human embryonic kidney cells have been reported (Klein et al. 2001). In addition, Barton et al. (2004) also suggested that Con G targets NR1/NR2A/NR2B receptors, whereas CI-1041, a NR2B-specific antagonist, does not. Therefore it is possible that Con G inhibits NMDA receptor–mediated sEPSCs and NMDA-gated currents in cortical neurons through action on both diheteromeric (NR1/NR2B) and triheteromeric (NR1/NR2A/NR2B) NMDA receptors. The age-dependent differences in the action of Con G observed in this study may also explain the variability of Con G–induced block of NMDA-evoked currents in rat cortical neurons (43–85%) that was previously reported (Donevan and McCabe 2000).

Con G is a unique antagonist in that it attenuates the amplitude as well as the decay time constant of NMDA-mediated EPSCs in hippocampal brain slices obtained from adult rats (Barton et al. 2004). The mechanism underlying the decrease in decay time constants of NMDA-mediated EPSCs is currently unknown. At least two hypotheses concerning the effect of Con G on the decay time constant have been suggested. 1) It is possible that Con G, which contains a glycine in position 2, could interfere with the glycine co-agonist binding site on the NR1 subunit. This interaction might then decrease the decay time constant as the affinity for glutamate might be reduced for NMDA receptors in the presence of Con G. Such allosteric interactions between the Con G and glycine binding sites have been reported (Donevan and McCabe 2000; Hammerland et al. 1992; Mena et al. 1990). 2) The second hypothesis is that Con G, unlike CI-1041 or CP101606, binds to and inhibits current flow through any receptor containing at least one NR2B subunit. Therefore at the adult CA1–schaffer collateral synapse, Con G might block heterotrimeric receptors containing NR1, NR2A, and NR2B subunits (whereas CI-1041 does not) and would not bind to heterodimeric receptors containing only NR1/NR2A subunits. The kinetics of NR2A containing receptors are substantially faster than those receptors containing NR2B subunits and Con G, by blocking heterotrimeric receptors effectively, would seem to speed up the decay time constant of the residual ESPC (Barton et al. 2004).

The first of these two hypotheses was tested in this set of experiments. It was hypothesized that if there was an interaction with the glycine-binding site, the dose-dependent inhibition of NMDA mediated EPSCs by Con G (0.01–0.3 μM) in cultures at 13 DIV at high and low concentrations of glycine would have been different (Fig. 3). In fact, similar inhibition of sEPSCs, as well as agonist-evoked currents, was observed at varying concentrations of glycine at 13 DIV (Fig. 4). The glycine dependence of Con G action in cultures from 16 and 19 DIV was most likely caused by an increase in the expression of NR2A subunits in these neurons as they age. The NR2A subunit has a 10-fold lower affinity for glycine than the NR2B subunit (Dingledine et al. 1999; Kutsuwada et al. 1992; Priestley et al. 1995), so that under low glycine conditions, the majority of receptors activated contained primarily NR2B subunits, and hence Con G seemed to have an increased efficacy. Greater inhibition of NMDA currents by ifenprodil, an NR2B-specific antagonist, has been shown previously in hippocampal neurons in the presence of low glycine concentrations, suggesting that different subtypes of NMDA receptors were selectively activated in the presence of high or low glycine concentrations (Kendrick et al. 1998). These data on the Con G–induced inhibition of NMDA currents at varying concentrations of glycine is in agreement with the earlier observation that Con G does not act directly at the glycine site (Donevan and McCabe 2000).

In conclusion, the data presented herein provide evidence that Con G–induced inhibition of synaptic and extrasynaptic NMDA receptors is most likely not caused by any modulation by Con G at the glycine binding site. The Con G–induced inhibition of NMDA receptors was both concentration and age-dependent. The ability of Con G to inhibit whole cell currents and sEPSCs at 19 DIV indicates its probable action at NR2A-2B containing receptors. Our results in cultured neurons are also consistent with Con G–induced inhibition of NMDA-mediated EPSCs seen in adult hippocampal brain slices (Barton et al. 2004). In addition, the decreased efficacy of Con G corresponding with the developmental switch in NR2 subunits in cortical neurons in vitro suggests that this model could be used effectively to study the structure–function relationship of conantokins with NMDA receptors that are developmentally regulated.

GRANTS

This work was supported by National Institute of General Medical Sciences Grant PO1 GM-048677-11 to B. M. Olivera and K. S. Wilcox.

Acknowledgments

The authors thank Dr. Annette E. Fleckenstein for providing laboratory space and support for doing the Western blot studies and Drs. Peter J. West, Evan L. Riddle, and Baldomero M. Olivera for thoughtful comments.

Footnotes

  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

REFERENCES

View Abstract