Decreased sensitivity to Group III mGluR agonists in the lateral perforant path following kindling

Neuropharmacology 38 (1999) 927 – 933

Decreased sensitivity to Group III mGluR agonists in the lateral perforant path following kindling Gloria J. Klapstein a, Brian S. Meldrum b, Istvan Mody a,* a

Department of Neurology, UCLA School of Medicine, 710 Westwood Plaza, Los Angeles, CA 90095 -1769, USA b Department of Clinical Neuroscience, Institute of Psychiatry, King’s College, London, UK Accepted 14 January 1999

Abstract The ability of the selective Group III mGluR agonist L-serine-O-phosphate (L-SOP) to inhibit lateral perforant path (LPP) evoked responses in the dentate gyrus was tested in hippocampal slices from commissurally-kindled rats 1 – 2 days after the last seizure, implanted controls, and fully-kindled rats rested for 28 days without stimulated seizures (28 days post-seizure, 28 dps). L-SOP was more potent in controls than kindled or 28 dps animals, decreasing the fEPSP slope with IC50s of 2.4 mM, 18.7 mM and 10.5 mM, respectively. Paired pulse facilitation (PPF, 50 ms) was comparable in control and kindled rats, but was markedly reduced in 28 dps rats, indicating increased release probability. Inhibition of the field excitatory postsynaptic potentials (fEPSP) by L-SOP was correlated with enhanced PPF in all groups, affirming a presynaptic site of action. At moderate levels of L-SOP-induced inhibition (20–60%), PPF showed significantly greater enhancement in 28 dps than in the other two groups. These results are interpreted as showing a functional reduction of the presynaptic inhibitory Group III mGluR (probably mGluR8) response in the LPP after kindling. Furthermore, PPF changes indicate that the kindled state may be associated with a long-lasting increase in the probability of release from LPP terminals, which may be temporarily masked or counterbalanced by recent seizures. © 1999 Elsevier Science Ltd. All rights reserved. Keywords: Epilepsy; mGluR7; mGluR8; Release probability; Paired pulse facilitation; Rat

1. Introduction Electrical kindling in the rat provides a model for complex partial seizures in man. In particular it has enhanced our understanding of the molecular and cellular changes that occur during the process of epileptogenesis (Mody, 1993; McNamara, 1995). Studies of dentate granule cells have revealed changes in the functional properties of NMDA and GABAA receptors and in voltage-sensitive calcium channels that could contribute to the fully-kindled state (Mody, 1998). Changes in the functional properties of metabotropic glutamate receptors (mGluR) have also been described. Phosphoinositide hydrolysis in response to quisqualate or (1S,3R)-1-aminocyclopentane-1,3-dicarboxylic acid (1S,3R ACPD) is enhanced either transiently or for 1 * Corresponding author. Tel.: +1 310 2063484; fax: + 1 310 8250033. E-mail address: [email protected] (I. Mody)

month or more following amygdala kindling (Akiyama et al., 1992), possibly leading to a sustained enhancement of protein kinase C activity in the kindled hippocampus (Akiyama et al., 1995). Electrophysiological studies in amygdala slices from amygdala-kindled rats indicate that the depolarizing action of Group I agonists is potentiated (Holmes et al., 1996). The presynaptic actions mediated via Group II and Group III mGluRs in basolateral amygdala neurons also appear to be potentiated (28–30-fold) in the fully-kindled brain (Neugebauer et al., 1997). Group III agonists such as L-2-amino-4-phosphonobutyric acid (L-AP4) and L-serine-O-phosphate (LSOP) act presynaptically to inhibit synaptic transmission evoked by lateral perforant path stimulation (Koerner and Cotman, 1981; Johansen et al., 1995; Bushell et al., 1996; Macek et al., 1996). Functional changes in these receptors would, therefore, alter epileptiform activity progressing through this pathway. We have sought to examine changes in the functional

0028-3908/99/$ - see front matter © 1999 Elsevier Science Ltd. All rights reserved. PII: S 0 0 2 8 - 3 9 0 8 ( 9 9 ) 0 0 0 1 6 - 7

928

G.J. Klapstein et al. / Neuropharmacology 38 (1999) 927–933

effect of presynaptic Group III mGluRs associated with hippocampal kindling. We have recorded dentate granule cell dendritic field excitatory postsynaptic potentials (fEPSPs) in response to stimulation of the lateral perforant path in hippocampal slices from commissurallykindled rats and control implanted rats. Using the paired pulse facilitation ratio (the ratio of the slopes of two fEPSPs evoked 50 ms apart) for both identification of lateral perforant path responses (McNaughton, 1980) and as an index of synaptic release probability at these synapses, we examined the concentration–response relationship for L-SOP. We have used the Group III selective antagonist (RS)-a-methylserine-Ophosphate (MSOP) (Jane et al., 1994; Thomas et al., 1996) to ensure the specificity of action of L-SOP.

subset of these rats were maintained without stimuli for at least 28 days following full kindling (28 day postseizure (dps) group), to serve as a control for the effects of recent seizure. Rats used as controls were also implanted, but received no stimuli.

2.2. Slice preparation

2. Methods

Rats were decapitated under sodium pentobarbital (i.p., 75 mg/kg) anesthesia. Brains were removed into ice-cold carbogenated (95% O2, 5% CO2) artificial cerebrospinal fluid (ACSF; in mM: 126 NaCl, 2 MgCl2, 2 CaCl2, 2.5 KCl, 1.25 NaH2PO4, 26 NaHCO3, 10 glucose), and sliced horizontally at 500 mm with a Vibratome. Slices were maintained in an incubation chamber filled with carbogenated ACSF at 32°C for at least 1 h before being transferred to the recording chamber.

2.1. Kindling

2.3. Stimulation and recording

Male Wistar rats (250 – 350 g) were stereotaxically implanted under sodium pentobarbital anaesthesia (i.p., 75 mg/kg), with bipolar stainless steel electrodes (Plastics One, Inc., Roanoke, VA) in the midline hippocampal commissure (AP −1.8; L 0.0; V 4.2) (Ko¨hr and Mody, 1991). Following 1 week of recovery, rats were stimulated five times per week with trains of stimuli (60 Hz for 1 s, 150 mA) and monitored for seizures. Only rats exhibiting at least 15 consecutive stage 5 seizures (Racine, 1972) were used in subsequent studies. A

In the recording chamber, slices were perfused constantly with carbogenated ACSF (34–35°C; composition as above, but with 50 mM picrotoxin to block GABAA responses, and an additional 2 mM CaCl2 and 2 mM MgCl2 to prevent the occurrence of spontaneous bursting) in an atmosphere of warm, moist carbogen. A bipolar stimulating electrode was placed in the outer third of the molecular layer of the upper blade of the dentate gyrus, in order to preferentially stimulate lateral perforant path fibers. Pairs of stimuli (100–200 ms,

Fig. 1. Under control conditions, paired pulse facilitation ratios from the 28 dps group were significantly lower than those observed in the other two groups (one-way ANOVA, P(F) = 0.04, Duncan’s post-hoc test). Points represent the ratios of the slopes of two fEPSPs elicited 50 ms apart (see inset, taken from control group) in slices prior to any addition of mGluR agonist. Means9S.E.M. for each group are represented by bar graphs.

G.J. Klapstein et al. / Neuropharmacology 38 (1999) 927–933

929

Fig. 2. The Group III mGluR agonist L-SOP inhibits lateral perforant path-evoked fEPSPs in the dentate gyrus. Upper panel: time course of inhibition and recovery following a 10 min application of 10 mM L-SOP (bar). Points represent means9 S.E.M. of normalized fEPSP slopes. Lower panel: representative traces showing inhibition of the slope of the fEPSP by L-SOP. Inhibition is reduced in kindled preparations, and remains reduced in the absence of recent seizures (28 dps group).

10 – 40 V, 50 ms inter-stimulus interval) were delivered every 30 s. Stimulus intensity was chosen such that the first stimulus produced a response 50 – 70% of maximum. Dendritic field EPSPs were recorded with glass microelectrodes (2– 5 MV) containing ACSF, also placed in the outer third of the molecular layer. Paired pulse facilitation was used as a criterion for proper electrode placement (McNaughton, 1980; Macek et al., 1996), and care was taken to find responses with PPF in all three groups. PPF was expressed as the ratio of the slope of the second EPSP to the slope of the first EPSP following a pair of stimuli. Recordings exhibiting paired pulse depression were rejected. A 10 mM stock solution of L-SOP in equimolar NaOH was made, and was diluted in ACSF before being bath applied.

2.4. Drugs L-serine-O-phosphate and (R,S)-a-methylserine-Ophosphate were from Tocris Ltd., UK; picrotoxin from Sigma Chemical Co., USA. All others were from Fluka Chemical Corp., USA.

2.5. Data analysis Data were digitized at 5 kHz, captured using the program SCAN (Strathclyde Electrophysiological Software, J. Dempster), and analyzed offline using the program Synapse (Y. de Koninck) to detect and measure the peak initial slope of the fEPSP. Non-linear

regression analysis and F-tests were performed on concentration–response data using the SAS statistical program, using the Hill equation I= Imax(x n/(k n +x n)), where Imax is the maximum inhibition produced, k is the IC50, x is concentration, and n is the Hill coefficient for binding of the agonist to the receptor. All other data were compared using one-way ANOVA in the SPSS statistical program. Data are presented as mean9 S.E.M. Statistical significance is set at P5 0.05.

3. Results There were a total of 78 drug applications made to 17 slice preparations from five sham-implanted control, seven kindled (1–2 dps), and five 28 dps animals. Initial placement of both stimulation and recording electrodes in the outer third of the molecular layer of the dentate gyrus readily yielded responses with PPF in slices from control (1.339 0.07, n= 5) and kindled (1.3290.08, n=7) groups. In the 28 dps group, however, PPF was much less robust (1.08 9 0.02, n=5; P(F) B 0.05, one-way ANOVA; Fig. 1) and more difficult to find, usually requiring that the recording and/or stimulation electrodes be moved several times within the outer third of the molecular layer before PPF was evident at all. L-SOP (1–300 mM) reversibly reduced the initial slope of the lateral perforant path–dentate granule cell field EPSP (Fig. 2) in a concentration-dependent man-

930

G.J. Klapstein et al. / Neuropharmacology 38 (1999) 927–933

ner (Fig. 3). In control preparations, L-SOP inhibited the LPP evoked response with an IC50 of 2.49 0.2 mM, a maximal effect of 59.3 91.5%, and a Hill coefficient of 1.1790.15. Co-application of the Group III mGluR antagonist MSOP (200 mM) in the presence of 30 mM L-SOP reversed the L-SOP-induced inhibition by 75% (data not shown). Kindling was associated with a significant decrease in L-SOP potency. In kindled preparations, the IC50 was significantly increased to 18.792.8 mM (P(F) B0.05); however, neither the maximum effect (69.5 91.5%) nor the Hill coefficient (0.8790.07) were significantly changed from control values (P(F) \0.125; Fig. 2, Table 1). L-SOP potency remained decreased in the 28 dps group, with an IC50 of 10.5 9 2.7 mM, which was significantly different from the control (P(F) B0.05), but not the kindled group (P(F) \0.125); neither the maximum effect (64.99 4.7%) nor the Hill coefficient (0.98 90.17) were significantly different from the other two groups. L-SOP-induced inhibition of the first fEPSP was accompanied by an increase in the PPF, consistent with a presynaptic site of action. Since synapses with low probability of release show relatively greater paired pulse facilitation (Dobrunz and Stevens, 1997), we used the L-SOP-induced presynaptic inhibition (i.e. the reduction in the probability of release) to examine the possible effects of kindling on glutamate release probability in this pathway. If the probability of release was the same in all groups, then an identical presynaptically-mediated reduction in the probability, such as that caused by L-SOP, should result in an identical shift in the PPF ratio. Data were binned by the amount of

inhibition of the first response, in increments of 20% (independent of L-SOP concentration), and the concomitant increase in PPF was compared between groups. Within each bin, the average inhibition did not differ between test groups (one-way ANOVA, P(F) \ 0.25, all bins). The PPF increase during L-SOP inhibition was similar in kindled and control preparations, except for a small but significant difference during a moderate degree of inhibition (20–40%), with less increase in PPF in kindled animals than in controls (P(F) B 0.05, one-way ANOVA, Duncan’s post-hoc test). In contrast, PPF in the 28 dps group was significantly greater than in both other groups during L-SOP induced inhibitions of 20–60% (P(F) B 0.05, Fig. 4).

4. Discussion In the dentate gyrus, the lateral perforant path terminates predominantly in the outer one-third of the molecular layer. It shows paired pulse facilitation and the excitatory responses are depressed by Group III mGluR agonists L-AP4 and L-SOP (McNaughton, 1980; Koerner and Cotman, 1981; Bushell et al., 1996; Macek et al., 1996). We have found that commissural kindling is associated with a long-lasting decrease in the potency of the Group III mGluR agonist L-SOP as judged by the reduction of the initial slope of the dentate granule cell dendritic field EPSP following LPP stimulation. By contrast, the potency of agonists of Group II and Group III metabotropic receptors in the basolateral amygdala of amygdala-kindled rats is increased 28–30-

Fig. 3. Concentration– response curves for inhibition of the fEPSP slope by L-SOP. Points represent means 9 S.E.M. Kindling is associated with a 10-fold decrease in potency of this agonist.

G.J. Klapstein et al. / Neuropharmacology 38 (1999) 927–933

931

Table 1 Effect of kindling on L-SOP potency

Control Kindled 28 dps

Max effect (% 9S.E.M.)

IC50 (mM9S.E.M.)

Hill ( 9 S.E.M.)

59.39 1.5 69.5 9 3.2 64.99 4.7

2.4 90.2 18.79 2.8 10.5 9 2.7

1.179 0.15 0.8790.07 0.989 0.17

fold as judged by reduction of EPSPs induced by stimulation of the lateral amygdala (Neugebauer et al., 1997), indicating that the effect of kindling on mGluR sensitivity is not consistent throughout the brain. However, such an impaired sensitivity of presynaptic receptors which decrease glutamate release could contribute to the enhanced glutamate release seen in the hippocampus at the onset of limbic seizures in patients (During and Spencer, 1993). In a parallel study with paired pulse stimulation of the perforant path in hippocampal slices from anterior temporal lobectomy specimens from patients with drug-resistant complex partial seizures, Kral et al. (1997) found that L-AP4 depressed the EPSP slope in patients without Ammon’s horn sclerosis but not in those with Ammon’s horn sclerosis and mossy fibre sprouting. Whether there is a causal relationship between the synaptic reorganization and changes in mGluR sensitivity is, as yet, unknown. In the present study, the similarity in L-SOP efficacy seen in all test groups does not suggest a change in receptor number. Rather a prolonged exposure to high levels of glutamate may have caused desensitization of the mGluRs. For example, it has been shown that mGluR5 undergoes a rapid phosphorylation-mediated desensitization upon a brief exposure to glutamate (Gereau and Heinemann, 1998). Although this effect may be too transient (i.e. recovering within 30 min) to account for our observation of decreased agonist potency at 28 days following a seizure, it is conceivable that a more permanent form of mGluR desensitization may exist in the kindled hippocampus. Which of the Group III mGluRs are involved in this modification of lateral perforant path responses? Electron microscopy reveals mGluR4a, mGluR7a, mGluR7b and mGluR8 located presynaptically on asymmetric and to a lesser extent on symmetric synapses in the dentate molecular layer (Shigemoto et al., 1997). Immunohistochemistry combined with lesioning of the perforant path shows that the lateral perforant path terminals in the outer molecular layer of the dentate gyrus are strongly reactive for mGluR8. By contrast, mGluR2 and mGluR7a are prominently expressed in the regions of termination of the medial perforant path where evoked responses are preferentially depressed by Group II agonists (Bushell

et al., 1996; Macek et al., 1996; Dietrich et al., 1997). Studies on the agonist sensitivity of mGluR7 suggest that it is unlikely to be contributing to the response we observe. Cloned rat mGluR7, when expressed in Chinese hamster ovary cells, reacts with L-AP4 and L-SOP to inhibit forskolin-induced cyclic AMP accumulation with an IC50 of 160 mM (for both agonists, Okamoto et al., 1994). In contrast, rat mGluR8 shows an IC50 of 0.4 mM for L-AP4 (Duvoisin et al., 1995) which is more compatible with the IC50 of 2.4 mM for L-SOP in our preparation. While it may be argued that the IC50s for the two agonists should not be directly compared, it should be noted that their inhibitory effects on forskolin stimulated cAMP formation in the hippocampus are not significantly different (Wright and Schoepp, 1996), and that similarly, the suppression of forskolin induced cAMP formation in CHO cells expressing mGluR7 is identical (Okamoto et al., 1994). It is, therefore, most probable that our results reflect a reduced mGluR8 response following kindling. The Group III mGluRs are thought to reduce the probability of glutamate release by decreasing Ca2 + entry via N-type and P/Q-type Ca2 + channels (Herlitze et al., 1997). This is consistent with our observation that L-SOP induced inhibition of LPP responses was accompanied by an increase in paired pulse facilitation (Dobrunz and Stevens, 1997). However, the relationship between presynaptic inhibition by L-SOP and the change in PPF was not constant between the different groups of rats tested. It was considerably more difficult to find PPF in 28 dps animals than in either control animals or kindled animals which had experienced recent seizures. Furthermore, at some moderate levels of L-SOP-induced inhibition in the three groups, kindled animals with or without recent seizures showed significantly less or more increase in PPF, respectively, compared to controls. In the outer third of the molecular layer of 28 dps animals, there may be more contamination from terminals of the reorganized medial perforant path, which exhibits paired pulse depression (Macek et al., 1996). Alternatively, the kindled state may be coincident with an increased probability of release at LPP terminals. The former possibility is unlikely, because L-SOP at the concentrations used selectively affects LPP synapses and its maximum effect is similar in all

932

G.J. Klapstein et al. / Neuropharmacology 38 (1999) 927–933

Fig. 4. L-SOP induced reductions in fEPSP slope were associated with increases in the paired pulse facilitation, consistent with a presynaptic site of action involving a decrease in the probability of glutamate release. Plot shows data binned by percentage inhibition of the first fEPSP slope. Within each bin, the average inhibition did not differ between test groups (one-way ANOVA, P(F) \0.21). During moderate degrees of presynaptic inhibition (20 – 60%), 28 dps animals showed a significantly greater concomitant increase in paired pulse facilitation than did the other two test groups (one-way ANOVA, Duncan’s multiple range post-hoc test). At 20 – 40% levels of inhibition by L-SOP, kindled animals demonstrated significantly smaller increases in PPF than the other two groups. Statistically significant differences are indicated by asterisks.

three groups. Either way, however, one would have to presume an opposing effect (i.e. a transient decrease in the probability of release) produced by recent seizures to explain the difference between the kindled and 28 dps groups. Many transient changes are known to occur following seizures — notably, changes in expression of neuropeptides (Rosen et al., 1992; Schwarzer et al., 1996) and their receptors (Crain et al., 1987). A net effect on presynaptic inhibition by any of these mechanisms would result in the decrease in LPP release probability which is consistent with our observations. Due to the extremely dynamic nature of neuronal communication following seizure activity, it is difficult to accurately predict the functional consequences of a decrease in presynaptic mGluR-mediated inhibition. While it seems intuitive that such a change would result in a net increase in glutamate release at affected synapses, it is also possible for the enhanced probability of release to attenuate high frequency transmission (Brenowitz et al., 1998), through the LPP.

Acknowledgements This work is supported by NIH grant NS-36142 to I.M. GJK holds a postdoctoral fellowship from the Alberta Heritage Foundation for Medical Research. Special thanks to K. Klapstein for assistance with non-linear regression analysis.

References Akiyama, K., Daigen, A., Yamada, N., Itoh, T., Kohira, I., Ujike, H., Otsuki, S., 1992. Long-lasting enhancement of metabotropic excitatory amino acid receptor-mediated polyphosphoinositide hydrolysis in the amygdala/pyriform cortex of deep prepiriform cortical kindled rats. Brain Res. 569, 71 – 77. Akiyama, K., Ono, M., Kohira, I., Daigen, A., Ishihara, T., Kuroda, S., 1995. Long-lasting increase in protein kinase C activity in the hippocampus of amygdala-kindled rat. Brain Res. 679, 212–220. Brenowitz, S., David, J., Trussell, L., 1998. Enhancement of synaptic efficacy by presynaptic GABAB receptors. Neuron 20, 135–141. Bushell, T.J., Jane, D.E., Tse, H.W., Watkins, J.C., Garthwaite, J., Collingridge, G.L., 1996. Pharmacological antagonism of the actions of group II and III mGluR agonists in the lateral perforant path of rat hippocampal slices. Br. J. Pharmacol. 117, 1457 – 1462. Crain, B.J., Chang, K.J., McNamara, J.O., 1987. An in vitro autoradiographic analysis of mu and delta opioid binding in the hippocampal formation of kindled rats. Brain Res. 412, 343–351. Dietrich, D., Beck, H., Kral, T., Clusmann, H., Elger, C.E., Schramm, J., 1997. Metabotropic glutamate receptors modulate synaptic transmission in the perforant path: pharmacology and localization of two distinct receptors. Brain Res. 767, 220–227. Dobrunz, L.E., Stevens, C.F., 1997. Heterogeneity of release probability, facilitation, and depletion at central synapses. Neuron 18, 995 – 1008. Duvoisin, R.M., Zhang, C., Ramonell, K., 1995. A novel metabotropic glutamate receptor expressed in the retina and olfactory bulb. J. Neurosci. 15, 3075 – 3083. During, M.J., Spencer, D.D., 1993. Extracellular hippocampal glutamate and spontaneous seizure in the conscious human brain. Lancet 341, 1607 – 1610.

G.J. Klapstein et al. / Neuropharmacology 38 (1999) 927–933 Gereau, R.W., Heinemann, S.F., 1998. Role of protein kinase C phosphorylation in rapid desensitization of metabotropic glutamate receptor 5. Neuron 20, 143–151. Herlitze, S., Hockerman, G.H., Scheuer, T., Catterall, W.A., 1997. Molecular determinants of inactivation and G protein modulation in the intracellular loop connectin domains I and II of the calcium channel a1A subunit. Proc. Natl. Acad. Sci. USA 94, 1512 – 1516. Holmes, K.H., Keele, N.B., Shinnick-Gallagher, P., 1996. Loss of mGluR-mediated hyperpolarizations and increase in mGluR depolarizations in basolateral amygdala neurons in kindling-induced epilepsy. J. Neurophysiol. 76, 2808–2812. Jane, D.E., Jones, P.L.St.J., Pook, P.C.-K., Tse, H.-W., Watkins, J.C., 1994. Actions of two new antagonists showing selectivity for different sub-types of metabotropic glutamate receptor in the neonatal rat spinal cord. Br. J. Pharmacol. 112, 809–816. Johansen, P.A., Chase, L.A., Sinor, A.D., Koerner, J.F., Johnson, R.L., Robinson, M.B., 1995. Type 4a metabotropic glutamate receptor: identification of new potent agonists and differentiation from the L-(+ )-2-amino-4-phosphonobutanoic acid-sensitive receptor in the lateral perforant pathway in rats. Mol. Pharmacol. 48, 140 – 149. Koerner, J.F., Cotman, C.W., 1981. Micromolar L-2-amino-4-phosphonobutyric acid selectively inhibits perforant path synapses from lateral entorhinal cortex. Brain Res. 216, 192–198. Ko¨hr, G., Mody, I., 1991. Endogenous intracellular calcium buffering and the activation/inactivation of HVA calcium currents in rat denate gyrus granule cells. J. Gen. Physiol. 98, 941–967. Kral, T., Dietrich, D., Clusmann, H., Schramm, J., 1997. Mossy fiber sprouting correlates with loss of synaptic depression by Group III metabotropic glutamate receptors (mGluRs) in the area dentata of patients with temporal lobe epilepsy (TLE). Soc. Neurosci. Abstr. 23, 817 (abstract). Macek, T.A., Winder, D.G., Gereau, R.W., Ladd, C.O., Conn, P.J., 1996. Differential involvement of Group II and Group III mGluRs as autoreceptors at lateral and medial perforant path synapses. J. Neurophysiol. 76, 3798–3806. McNamara, J.O., 1995. Analyses of the molecular basis of kindling development. Psychiatry Clin. Neurosci. 49, S175–S178.

933

McNaughton, B.L., 1980. Evidence for two physiologically distinct perforant pathways to the fascia dentata. Brain Res. 199, 1–19. Mody, I., 1993. The molecular basis of kindling. Brain Pathol. 3, 395 – 403. Mody, I., 1998. Ion channels in epilepsy. Int. Rev. Neurobiol. 42, 199 – 226. Neugebauer, V., Keele, N.B., Shinnick-Gallagher, P., 1997. Epileptogenesis in vivo enhances the sensitivity of inhibitory presynaptic metabotropic glutamate receptors in basolateral amygdala neurons in vitro. J. Neurosci. 17, 983 – 995. Okamoto, N., Hori, S., Akazawa, C., Hayashi, Y., Shigemoto, R., Mizuno, N., Nakanishi, S., 1994. Molecular characterization of a new metabotropic glutamate receptor mGluR7 coupled to inhibitory cyclic AMP signal transduction. J Biol Chem. 269, 1231– 1236. Racine, R.J., 1972. Modification of seizure activity by electrical stimulation II. Motor seizure. Electroencephalogr. Clin. Neurophysiol. 32, 281 – 294. Rosen, J.B., Cain, C.J., Weiss, S.R., Post, R.M., 1992. Alterations in mRNA of enkephalin, dynorphin and thyrotropin releasing hormone during amygdala kindling: an in situ hybridization study. Mol. Brain Res. 15, 247 – 255. Schwarzer, C., Sperk, G., Samanin, R., Rizzi, M., Gariboldi, M., Vezzani, A., 1996. Neuropeptides-immunoreactivity and their mRNA expression in kindling: functional implications for limbic epileptogenesis. Brain Res. Rev. 22, 27 – 50. Shigemoto, R., Kinoshita, A., Wada, E., Nomura, S., Ohishi, H., Takada, M., Flor, P.J., Neki, A., Abe, T., Nakanishi, S., Mizuno, N., 1997. Differential presynaptic localization of metabotropic glutamate receptor subtypes in the rat hippocampus. J. Neurosci. 17, 7503 – 7522. Thomas, N.K., Jane, D.E., Tse, H.W., Watkins, J.C., 1996. a-methyl derivatives of serine-O-phosphate as novel, selective competitive metabotropic glutamate receptor antagonists. Neuropharmacology 35, 637 – 642. Wright, R.A., Schoepp, D.D., 1996. Differentiation of group 2 and group 3 metabotropic glutamate receptor cAMP responses in the rat hippocampus. Eur. J. Pharmacol. 297, 275 – 282.

. .