INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Transient activation of group I metabotropic glutamate receptors (mGluRs) converts interictal bursts into seizure-length discharges (Merlin 1999; Merlin and Wong 1997), and in vivo studies reveal that selective group I mGluR agonists are convulsants (Camón et al. 1998; Thomsen and Dalby 1998). Moreover, group I mGluR-induced burst prolongation persists for hours following washout of the selective agonist, and group I mGluR antagonists reversibly suppress the persistent prolonged bursts (Merlin 1999; Merlin and Wong 1997). These findings implicate group I mGluRs not only in the expression of seizures but in the induction and maintenance of epileptogenesis, the process in which the neuronal cortical network becomes persistently predisposed to the production of seizure discharges.

The group I mGluRs are comprised of two receptor subtypes (mGluR1 and mGluR5), both of which are present in the hippocampal CA3 region (Blümcke et al. 1996; Fotuhi et al. 1994; Luján et al. 1996). Although both subtypes are linked to inositol-1,4,5-triphosphate (IP3) generation and calcium mobilization (Abe et al. 1992; Houamed et al. 1991; Masu et al. 1991), mGluR1 and mGluR5 generate different intracellular calcium responses (Kawabata et al. 1998). Furthermore, in kindling as well as human temporal lobe epilepsy studies, mGluR1 immunoreactivity and mRNA expression are upregulated while mGluR5 mRNA is either unchanged or downregulated (Akbar et al. 1996; Al-Ghoul et al. 1998; Blümcke et al. 2000), demonstrating independent regulatory control of the two receptors. Thus mGluR1 and mGluR5 may respond differently to similar stimuli, resulting in different long-term effects. Studies described herein therefore examine the contributions of each receptor subtype to the induction and maintenance of the persistent potentiation of epileptiform activities in the hippocampal slice. Portions of this work have appeared in abstract form (Merlin 2000).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Guinea pigs 2-4 wk old were anesthetized with halothane and decapitated in accordance with standard protocols. The brain was placed in ice-cold artificial cerebrospinal fluid (ACSF) containing (in mM) 124 NaCl, 26 NaHCO₃, 5 KCl, 1.6 MgCl₂, 2 CaCl₂, and 10 D-glucose. Transverse hippocampal slices (400 µm) were prepared with a Vibratome (Technical Products International) and then transferred to nylon mesh in an interface recording chamber (Fine Science Tools). Slices were oxygenated and humidified with 95% O₂-5% CO₂ at 35-36°C, perfused with ACSF, and incubated at least 1 h prior to recording.

Recordings from CA3 stratum pyramidale used 30- to 80- or 2- to 5-M glass microelectrodes filled with 2 M potassium acetate or 2 M NaCl (for intracellular or extracellular recording, respectively). Data were amplified and digitized using an Axoclamp 2B and DigiData 1200 series interface (Axon Instruments) and analyzed using pClamp and SigmaPlot software. All chemicals were bath-applied. Picrotoxin (PTX, 50 µM), an antagonist of GABA_A receptor-mediated inhibition, was present throughout all experiments. (S)-3,5-dihydroxyphenylglycine (DHPG), (+)-2-methyl-4-carboxyphenylglycine (LY367385), and 2-methyl-6-(phenylethynyl)-pyridine (MPEP) were obtained from Tocris Cookson (Ballwin, MO); all other chemicals were from Sigma (St. Louis, MO).

Sample sizes (n) refer to number of slices tested. Burst frequency (Hz) represents the reciprocal of interburst interval (seconds). Net burst prolongation was determined by subtracting the mean burst length induced in PTX ("control") from the prolonged bursts induced by DHPG. Data are reported as means ± SE. Statistical significance was determined using the Student's t-test, paired when appropriate, using the baseline picrotoxin bursts in a given slice as its own control. Additionally, to compare across groups (LY367385 vs. MPEP), ANOVA with post-ANOVA Newman-Keuls multiple comparison test was performed. P < 0.05 was considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Contribution of mGluR1 vs. mGluR5 in the maintenance of DHPG-induced persistent prolonged bursts

Synchronized epileptiform discharges 320-525 ms in duration recurred regularly at 0.10-0.15 Hz in CA3 pyramidal cells in guinea pig hippocampal slices perfused with PTX. The addition of 50 µM DHPG (a group I mGluR agonist) (Ito et al. 1992) elicited a rapid-onset increase in burst frequency, reaching a maximum of 0.38 ± 0.02 Hz at approximately 10 min (n = 8) (also see Galoyan and Merlin 2000; Merlin et al. 1998). This was accompanied by a progressive increase in epileptiform burst duration that persisted on removal of the agonist (BD at 45 min DHPG, 1,631 ± 174 ms, n = 8; e.g., Fig. 1A). We have previously shown that the maintenance of DHPG-induced prolonged epileptiform bursts can be reversibly suppressed with group I mGluR antagonists (Merlin 1999; Merlin and Wong 1997). To determine whether this effect is primarily mediated by mGluR1 or mGluR5, experiments were performed using either the selective mGluR1 antagonist LY367385 (Clark et al. 1997) or MPEP, a selective antagonist for mGluR5 (Gasparini et al. 1999).

View larger version (24K): [in this window] [in a new window]   Fig. 1. Both metabotropic glutamate receptor (mGluR)-1 and mGluR5 participate in the maintenance of (S)-3,5-dihydroxyphenylglycine (DHPG)-induced burst prolongation. A: example of time course of modification of picrotoxin-induced epileptiform activities; each burst represented by  (burst duration; top) and  (burst frequency; bottom). B: extracellular CA3 recordings corresponding to times on graphs in A. DHPG, (+)-2-methyl-4-carboxyphenylglycine (LY367385), and 2-methyl-6-(phenylethynyl)-pyridine (MPEP) traces were taken at end of drug application period; WASH was just prior to initiation of MPEP application. Horizontal calibration bars: left, 2 s; right, 500 ms. In other experiments, the order of antagonist application was reversed without consequence to the results reported.

In the presence of LY367385 (20-25 µM), a nearly complete, reversible suppression of the prolonged bursts was seen, similar in appearance to that seen with the group I mGluR antagonists (+)--methyl-4-carboxyphenylglycine or (S)-4-carboxyphenylglycine (Figs. 1 and 2) (cf. Merlin and Wong 1997). LY367385 reduced the BD from 1,670 ± 201 to 508 ± 110, an 89% reduction of net prolongation over control length (n = 4; Fig. 2). In comparison, MPEP (25 µM) produced only a partial suppression of the prolonged persistent epileptiform bursts, with a significant reduction in mean BD (P < 0.05) but with a wide variability in burst length (e.g., Fig. 1). The mean BD just prior to MPEP application (at 30-60 min of DHPG washout) was 1,712 ± 233 ms; MPEP reduced the BD to 889 ± 148 ms, a 64% reduction of the net prolongation (n = 4; Fig. 2). ANOVA analysis revealed that the BD in MPEP was significantly different from those seen in LY367385 and the two sets of control picrotoxin bursts (P < 0.05).

View larger version (14K): [in this window] [in a new window]   Fig. 2. Selective mGluR1 antagonist completely suppresses persistent burst prolongation; mGluR5 antagonist is only partially effective. A: 2 sets of experiments, comparing LY367385 to MPEP in their efficacy at suppressing the DHPG-induced persistent prolonged epileptiform bursts (n = 4 in each case; only the 1st antagonist applied after DHPG washout was used for this analysis). DHPG bar: burst duration (BD) after 30-45 min DHPG application and at least 30-min washout. *, P < 0.05 compared with PTX control. Although MPEP produces a significant reduction in burst length, the bursts remain significantly longer than the control bursts in PTX (P < 0.05). B: examples of experiments in A, CA3 intracellular recordings. a, DHPG-induced persistent burst before and after application of LY367385; b, before and after MPEP.

Contribution of mGluR1 vs. mGluR5 in DHPG-mediated induction of burst prolongation

Additional experiments were performed in which DHPG was applied in the presence of either MPEP or LY367385 (25 µM each). In all slices tested in the presence of MPEP, the induction of burst prolongation was markedly suppressed. The control burst duration in PTX + MPEP was 349 ± 20 ms; at 45-55 min of DHPG in the presence of MPEP, 411 ± 49 ms; after 1 h washout of DHPG and MPEP, 382 ± 56 ms (P > 0.05, n = 10) (Fig. 3B, 1 and 2). Nevertheless, the increase in burst frequency normally seen at 10 min of DHPG application was unaffected by the presence of MPEP, reaching a maximum frequency of 0.47 ± 0.03 Hz (n = 10; Fig. 3B1, bottom graph). Furthermore, burst prolongation could be elicited on reintroduction of DHPG in the absence of MPEP (Fig. 3B1, top graph).

View larger version (33K): [in this window] [in a new window]   Fig. 3. Selective mGluR5 activation can induce persistent prolonged bursts; selective mGluR5 antagonist prevents induction of mGluR-mediated burst prolongation. A1 and B1: summary data for experiments performed in the presence of LY367385 or MPEP, respectively. Selective antagonist was applied at least 10 min prior to initiating DHPG perfusion and was continued for 15 min following completion of DHPG application. *P < 0.05; **P < 0.01. A1 represents only the 5 slices in which burst prolongation appeared; B1: n = 10. A2 and B2: representative examples of experiments in A1 and B1; continuous intracellular recordings from CA3 pyramidal cells after 50 min co-application of agonist and selective antagonist (left) and after 1 h washout of the agonist (right). LY, LY367385.

DHPG application in the presence of LY367385 similarly elicited a prompt increase in burst frequency (0.35 ± 0.01 Hz at 10 min, n = 8; Fig. 3A1), and prolonged bursts were not expressed during co-application of DHPG and LY367385 (BD 418 ± 32 ms at 50 min DHPG in the presence of LY367385; P > 0.05 relative to control BD; n = 8). However, although in three slices BD remained unchanged at 1 h washout of the two agents (455 ± 25 ms, n = 3, P > 0.05), in five of eight slices, burst prolongation gradually appeared on washout (964 ± 196 ms at 1 h washout, n = 5, P < 0.05; Fig. 3A, 1 and 2).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Potentiation of mGluR1-mediated responses underlies maintenance of prolonged bursts

The data reveal that both mGluR1 and mGluR5 activation contribute independently to maintaining the persistent prolonged synchronized discharges (see Figs. 1 and 2). However, mGluR1 antagonist is much more effective at suppressing the persistent prolonged bursts, restoring the discharges to their control length in picrotoxin. This suggests that glutamate acting at mGlu1 receptors sustains the prolonged bursts. The burst prolongation may therefore be mediated predominantly by potentiation of mGluR1-mediated responses, similar to that shown accompanying both kindling-induced seizures and temporal lobe epilepsy (Akbar et al. 1996; Al-Ghoul et al. 1998; Blümcke et al. 2000). The involvement of mGluR5-mediated responses is less impressive (Figs. 1 and 2); this too is consistent with kindling studies, some of which demonstrate potentiation of responses (Keele et al. 2000) while others demonstrate mGluR5 downregulation (Akbar et al. 1996) or lack of change (Blümcke et al. 2000).

Activation of mGluR5 plays a critical role in the induction of long-lasting burst prolongation

The potentiation of group I mGluR-mediated responses described in the preceding text is induced via transient group I mGluR activation. Although the mGluR5 antagonist MPEP at 25 µM was only moderately effective in suppressing the maintenance of the potentiated bursts (Figs. 1 and 2), the same concentration was markedly effective in preventing the induction of long-lasting prolonged bursts (Fig. 3). In contrast, LY367385, when used at the same concentration that completely and reversibly suppressed the maintenance of the potentiated bursts, failed in 68% of slices to prevent the induction of the persistent burst prolongation. The differing induction properties of mGluR1 and mGluR5 may result from their different intracellular calcium responses. The calcium response to mGluR1 activation is mediated by transient calcium release from intracellular stores followed by sustained calcium entry through receptor-operated channels, whereas mGluR5 elicits a prolonged calcium oscillation entirely mediated by release from intracellular stores (Kawabata et al. 1998). Perhaps a store-mediated oscillatory calcium response is more effective in inducing epileptogenesis, and may prove especially effective in potentiating mGluR1 responses via an as-yet undetermined mechanism.

Greenwood et al. (2000) report that suppression of mGluR1 expression (via antisense oligonucleotide injections) dramatically suppresses the rate of progression of kindled seizures, suggesting that one may be able to prevent epileptogenesis via reducing mGluR1 function. However, it may be that the reduced mGluR1 function only serves to suppress seizure expression. Indeed, prolongation of epileptiform bursts was not elicited when DHPG was transiently applied in the presence of mGluR1 antagonist. Nevertheless, prolonged bursts did appear after the mGluR1 antagonist was removed (Fig. 3), revealing that the induction process stimulated by mGluR5 activation was able to proceed unimpeded, despite the suppressed expression of the seizure discharges. Given the stronger dependence on mGluR5 activation for induction of persistent prolonged bursts (Fig. 3), antagonism of mGluR5 may be required to accomplish the task of truly preventing epileptogenesis. Selective mGluR5 antagonists should therefore be studied for their potential clinical usefulness as antiepileptogenic agents.