Brain Research 910 (2001) 205–207www.elsevier.com/ locate /bres

Short communication

Synaptically-released zinc inhibits N-methyl-D-aspartate receptoractivation at recurrent mossy fiber synapses

*´Peter Molnar, J. Victor NadlerDepartment of Pharmacology and Cancer Biology and Department of Neurobiology, Duke University Medical Center, Durham, NC 27710, USA

Accepted 20 June 2001

Abstract

Hippocampal slices from pilocarpine-treated rats were used to explore the effect of zinc released at mossy fiber synapses on dentategranule cells. Chelation of zinc enhanced the N-methyl-D-aspartate (NMDA) receptor-mediated component of the excitatory postsynapticcurrent (EPSC), but did not affect the AMPA/kainate receptor-mediated component. Its effect was detectable only at negative membranepotentials and was pathway specific. Thus corelease of zinc reduces the ability of glutamate to activate postsynaptic NMDA receptors.Through this action, zinc would be expected to attenuate granule cell epileptiform activity supported by the recurrent mossy fiberpathway. 2001 Elsevier Science B.V. All rights reserved.

Theme: Disorders of the nervous system

Topic: Epilepsy: basic mechanisms

Keywords: Zinc; Mossy fibers; Epilepsy; Granule cell; Hippocampus; NMDA receptor

In many patients with temporal lobe epilepsy, hip- mossy fiber synapses on CA3 pyramidal cells, even singlepocampal mossy fibers send collaterals back into the electrical stimuli release enough zinc to reduce the NMDAmolecular layer of the dentate gyrus [15]. Studies of receptor-mediated component of the EPSC [16]. The EPSCanimal models demonstrated that recurrent mossy fiber at mossy fiber–granule cell synapses also has an NMDAgrowth creates monosynaptic recurrent excitatory connect- receptor-mediated component [10,12]. Blockade of NMDAions among dentate granule cells that are normally very receptors attenuates granule cell epileptiform activitysparse [10,12,17]. Expansion of the recurrent mossy fiber supported by the recurrent mossy fiber pathway [14]. Thispathway reduces the threshold for granule cell synchroni- result suggests that NMDA receptors at mossy fiber–zation [8,14] and may thus enhance the participation of granule cell synapses play an important role in promotingthese cells in seizures. such activity. Thus we tested the possibility that zinc

The hippocampal mossy fibers probably contain more regulates NMDA receptor function at this site.zinc than any other pathway in the brain. Within the mossy Whole cell patch clamp recordings were made fromfiber bouton, zinc is sequestered in synaptic vesicles and dentate granule cells in hippocampal slices prepared from

21can be released in a Ca -dependent manner by electrical pilocarpine-treated rats. The procedures used have beenstimulation of the pathway [6]. We examined the possi- described in detail previously [10,12]. Electrophysiologicalbility that release of zinc disinhibits dentate granule cells studies were performed 10–30 weeks after pilocarpineby blocking GABA receptors [3,7] and found no evidence administration. Timm histochemistry confirmed the pres-A

to support this idea [11]. Alternatively, zinc could reduce ence of recurrent mossy fiber growth in all these animals.granule cell excitability by regulating the participation of Slices were transferred to a beaker of modified artificialNMDA receptors in recurrent mossy fiber transmission. At cerebrospinal fluid (mACSF), which contained (in mM)

122 NaCl, 25 NaHCO , 3.5 KCl, 1.8 CaCl , 1.2 MgCl ,3 2 2

10 D-glucose, pH 7.4, and oxygenated with 95% O /5%2*Corresponding author. Tel.: 11-919-684-5317; fax: 11-919-681-CO for 1.5 h. A slice was then transferred to a submer-8609. 2

E-mail address: nadle002@acpub.duke.edu (J.V. Nadler). sion-type recording chamber mounted on the stage of a

0006-8993/01/$ – see front matter 2001 Elsevier Science B.V. All rights reserved.PI I : S0006-8993( 01 )02720-2

´206 P. Molnar, J.V. Nadler / Brain Research 910 (2001) 205 –207

confocal microscope and superfused with mACSF at roomtemperature. The internal solution used in the patchelectrode contained (in mM) 120 cesium gluconate, 10HEPES, 2 MgATP, 1 EGTA, 5 creatine phosphate, 20units /ml creatine phosphokinase, 10 QX-314 (N-ethyllidocaine) chloride and 0.1 Alexa Fluor 488 hydrazide(Molecular Probes, Eugene, OR), pH 7.4 and 276 mosm.Experiments began approximately 20 min after achievingwhole cell access. Signals were filtered at 2 kHz, sampledat 10 kHz and stored to disk. Traces shown in the figuresare averages of three consecutive recordings made 30 or 60s apart. Grouped data are expressed as means6S.E.M. andstatistical comparisons utilized the Wilcoxon signed ranktest.

The effect of 1 mM calcium disodium EDTA(CaEDTA), a high affinity zinc chelator, was tested onsynaptic transmission at recurrent mossy fiber and per-forant path synapses. At this concentration, CaEDTA

21reduced [Mg ] by ,1%. Bath application of CaEDTA0

consistently and reversibly enhanced the NMDA receptor-mediated component of the recurrent mossy fiber EPSC(11964% of control, n58; P,0.01), but did not sig- Fig. 1. CaEDTA (1 mM) selectively enhanced the NMDA receptor-

mediated component of the EPSC evoked by stimulating the recurrentnificantly affect the NMDA receptor-mediated componentmossy fibers. (Top) The NMDA receptor-mediated component of theof the perforant path EPSC (10367% of control, n56;EPSC was isolated pharmacologically by recording in the presence of 30

P.0.1) (Fig. 1). In some studies, the NMDA receptor-mM bicuculline and 5 mM NBQX (2,3-dihydroxy-6-nitro-7-sulfamyl-

mediated component of the recurrent mossy fiber EPSC benzo(F)quinoxaline) at a holding potential of 220 mV. Each set ofwas recorded at alternating holding potentials of 230 and recordings shows averaged traces from a representative experiment.

Addition of 50 mM D-AP5 (D-2-amino-5-phosphonopropanoate) to the140 mV (Fig. 2). CaEDTA increased the peak amplitudesuperfusion medium essentially abolished these responses. C, control; W,of this response when the membrane potential of thewash; *, stimulus artifact. (Bottom) Mean values6S.E.M. for eight

recorded cell was clamped at 230 mV (13065% of experiments on the recurrent mossy fiber pathway (MF) and six experi-control, n59; P,0.01), but not when it was clamped at ments on the perforant path (PP). Baseline amplitude was the average140 mV (9568% of control; P.0.1). CaEDTA did not value obtained during the 6 min period before addition of CaEDTA to the

superfusion medium.significantly affect the AMPA/kainate receptor-mediatedcomponent of the recurrent mossy fiber EPSC. The peakamplitude of responses recorded at a holding potential of specific, in that zinc appears to have no such regulatory280 mV was unchanged (9666% of control, n54; P. function at either NMDA receptors in the perforant path or0.1). at AMPA/kainate receptors in the recurrent mossy fiber

Although CaEDTA can bind a number of heavy metal pathway. These results are also consistent with the studiescations with high affinity, chelation of zinc probably of Vogt et al. [16].accounts for its action in this study because zinc is the Our finding that the action of CaEDTA on the NMDApredominant heavy metal contained in and released by receptor-mediated EPSC could be detected only at negativehippocampal mossy fibers. CaEDTA enhanced the activa- membrane potentials implies that released zinc inhibitstion of postsynaptic NMDA receptors by glutamate re- NMDA receptor activation in large part by acting upon itsleased from the recurrent mossy fiber pathway. This low affinity, voltage-dependent binding sites within thefinding implies that zinc, when coreleased with glutamate, channel [4,13]. CaEDTA cannot chelate all the zinc that isnormally serves to depress NMDA receptor function. Even coreleased with glutamate at mossy fiber synapses, how-the quantity of zinc released by a single stimulus appears ever, due to its slow binding kinetics [16]. Thus releasedsufficient for this purpose. Some investigators have sug- zinc may have an additional inhibitory action at its highgested, based on data obtained with indirect methods of affinity, voltage-independent site outside the channel; itslow sensitivity, that high frequency stimulus trains are lack of effect on NMDA receptor-mediated synapticrequired to release zinc from the mossy fiber pathway currents at positive membrane potentials does not exclude[1,2]. In agreement with our results, however, Vogt et al. this possibility.[16] found that zinc released from this pathway by a single The apparent inability of released zinc to alter thestimulus inhibits NMDA receptor activation at the synapse postsynaptic currents mediated by AMPA or kainateon CA3 pyramidal cells. The effect of zinc on NMDA receptors at the mossy fiber–granule cell synapse meritsreceptors at recurrent mossy fiber synapses is rather further comment. Exogenously applied zinc has been

´P. Molnar, J.V. Nadler / Brain Research 910 (2001) 205 –207 207

Acknowledgements

We thank Ms K. Gorham for secretarial assistance. Thisresearch was supported by NIH grant NS 38108.

References

[1] L. Anikstejn, G. Charton, Y. Ben-Ari, Selective release of endogen-ous zinc from the hippocampal mossy fibers in situ, Brain Res.404 (1987) 58–64.

21[2] S.Y. Assaf, S.-H. Chung, Release of endogenous Zn from braintissue during activity, Nature 308 (1984) 734–735.

[3] E.H. Buhl, T.S. Otis, I. Mody, Zinc-induced collapse of augmentedinhibition by GABA in a temporal lobe epilepsy model, Science 271(1996) 369–373.

[4] N. Chen, A. Moshaver, L.A. Raymond, Differential sensitivity ofrecombinant N-methyl-D-aspartate receptor subtypes to zinc inhibi-tion, Mol. Pharmacol. 51 (1997) 1015–1023.

[5] J.C. Dreixler, J.P. Leonard, Subunit-specific enhancement of gluta-mate receptor responses by zinc, Mol. Brain Res. 22 (1994) 144–150.

[6] C.J. Fredrickson, G. Danscher, Zinc containing neurons in hip-Fig. 2. CaEDTA (1 mM) enhanced the NMDA receptor-mediatedpocampus and related CNS structures, Prog. Brain Res. 83 (1990)component of the recurrent mossy fiber EPSC in a voltage-dependent71–84.manner. (Top) Experiments were carried out as indicated in Fig. 1, except

[7] J.W. Gibbs, M.D. Shumate, D.A. Coulter, Differential epilepsy-that the holding potential alternated between 230 and 140 mV. (Bottom)associated alterations in postsynaptic GABA receptor function inAMean values6S.E.M. for nine experiments. Amplitudes were computeddentate granule and CA1 neurons, J. Neurophysiol. 77 (1997)from the average of the three traces recorded at, immediately before and1924–1938.immediately after the indicated time. See Fig. 2 for other details.

[8] J.L. Hardison, M.M. Okazaki, J.V. Nadler, Modest increase inextracellular potassium unmasks effect of recurrent mossy fiber

shown to enhance the channel function of some, but not growth, J. Neurophysiol. 84 (2000) 2380–2389.[9] D.D. Lin, A.S. Cohen, D.A. Coulter, Zinc-induced augmentation ofall, AMPA receptors [5]. In particular, exogenously applied

excitatory synaptic currents and glutamate receptor responses inzinc can enhance AMPA receptor-mediated synaptic cur-hippocampal CA3 neurons, J. Neurophysiol. 85 (2001) 1185–1196.rents at mossy fiber-CA3 pyramidal cell synapses [9]. Our

´[10] P. Molnar, J.V. Nadler, Mossy fiber-granule cell synapses in theresult suggests either that the AMPA receptors expressed at normal and epileptic rat dentate gyrus studied with minimal laserthe mossy fiber–granule cell synapse are of a zinc-insensi- photostimulation, J. Neurophysiol. 82 (1999) 1883–1894.

´[11] P. Molnar, J.V. Nadler, Lack of effect of mossy fiber-released zinc ontive type or that exogenously applied zinc has actions notgranule cell GABA receptors in the pilocarpine model of epilepsy,Ashared by released zinc, perhaps because the concen-J. Neurophysiol. 85 (2001) 1932–1940.trations of exogenous zinc typically used in these studies

´[12] M.M. Okazaki, P. Molnar, J.V. Nadler, Recurrent mossy fiberare supraphysiological. pathway in rat dentate gyrus: synaptic currents evoked in presence

The ability of zinc released by the recurrent mossy fiber and absence of seizure-induced growth, J. Neurophysiol. 81 (1999)1645–1660.pathway to limit the activation of postsynaptic NMDA

[13] P. Paoletti, P. Ascher, J. Neyton, High-affinity zinc inhibition ofreceptors has important implications for its role in seizureNMDA NR1–NR2A receptors, J. Neurosci. 17 (1997) 5711–5725.propagation through the dentate gyrus. The NMDA re-

[14] P.R. Patrylo, F.E. Dudek, Physiological unmasking of new gluta-ceptor-mediated component of the unitary EPSC accounts matergic pathways in the dentate gyrus of hippocampal slices fromfor only a small percentage of the total synaptic response kainate-induced epileptic rats, J. Neurophysiol. 79 (1998) 418–429.

[15] T. Sutula, G. Cascino, J. Cavazos, I. Parada, L. Ramirez, Mossy[10]. Corelease of zinc with glutamate inhibits NMDAfiber synaptic reorganization in the epileptic human temporal lobe,receptor activation and may account, at least in part, forAnn. Neurol. 26 (1989) 321–330.the small contribution NMDA receptors make. Further-

[16] K. Vogt, J. Mellor, G. Tong, R. Nicoll, The actions of synapticallymore, other investigators have suggested that release of released zinc at hippocampal mossy fiber synapses, Neuron 26zinc from recurrent mossy fibers enhances granule cell (2000) 187–196.

[17] J.-P. Wuarin, F.E. Dudek, Electrographic seizures and new recurrentexcitability through disinhibition [3,7]. Our results suggest,excitatory circuits in the dentate gyrus of hippocampal slices fromto the contrary, that zinc may serve as a brake onkainate-treated epileptic rat, J. Neurosci. 16 (1996) 4438–4448.epileptiform activity by opposing activation of NMDA

receptors. Further studies are needed to resolve this issue.