The recurrent mossy fiber pathway of the dentate gyrus expands dramatically in the epileptic brain and serves as a mechanism for synchronization of granule cell epileptiform activity. It has been suggested that this pathway also promotes epileptiform activity by inhibiting GABAA receptor function through release of zinc. Hippocampal slices from pilocarpine-treated rats were used to evaluate this hypothesis. The rats had developed status epilepticus after pilocarpine administration, followed by robust recurrent mossy fiber growth. The ability of exogenously applied zinc to depress GABAA receptor function in dentate granule cells depended on removal of polyvalent anions from the superfusion medium. Under these conditions, 200 μM zinc reduced the amplitude of the current evoked by applying muscimol to the proximal portion of the granule cell dendrite (23%). It also reduced the mean amplitude (31%) and frequency (36%) of miniature inhibitory postsynaptic currents. Nevertheless, repetitive mossy fiber stimulation (10 Hz for 1 s, 100 Hz for 1 s, or 10 Hz for 5 min) at maximal intensity did not affect GABAAreceptor-mediated currents evoked by photorelease of GABA onto the proximal portion of the dendrite, where recurrent mossy fiber synapses were located. These results could not be explained by stimulation-induced depletion of zinc from the recurrent mossy fiber boutons. Negative results were obtained even during exposure to conditions that promoted transmitter release and synchronized granule cell activity (6 mM [K+]o, nominally Mg2+-free medium, 33°C). These results suggest that zinc released from the recurrent mossy fiber pathway did not reach a concentration at postsynaptic GABAA receptors sufficient to inhibit agonist-evoked activation.
The dentate gyrus is considered to serve as a “gate” or “filter” that restricts the ability of synchronous activity projected from the entorhinal cortex to invade the hippocampus (Lothman et al. 1992; Stringer et al. 1989). In the living brain, granule cells of the dentate gyrus normally fire action potentials at low rates (0.1–1 Hz) (Jung and McNaughton 1993). Although dentate granule cells can generate cellular bursts, the bursts generally remain isolated (Pan and Stringer 1996). Only when excitatory input from the entorhinal cortex is sufficiently powerful to activate the granule cell population synchronously, that is, to produce “maximal dentate activation,” does epileptiform activity propagate into the hippocampus (Stringer and Lothman 1992). Synaptic interconnections among principal neurons serve as the anatomic substrate for synchronization of cellular bursts in area CA3 of the hippocampus (Knowles et al. 1987; Traub et al. 1987). Thus the virtual lack of such connections among dentate granule cells accounts, in part, for the difficulty in provoking epileptiform granule cell discharge. In many patients with temporal lobe epilepsy (Babb et al. 1991; Franck et al. 1995; Represa et al. 1989; Sutula et al. 1989) and in several animal models of epilepsy (Mello et al. 1993; Nadler et al. 1980; Sutula et al. 1988), however, dentate granule cells become extensively interconnected through axonal growth and synaptogenesis. This recurrent mossy fiber pathway supports the pathological synchronization of granule cell epileptiform activity (Cronin et al. 1992;Hardison et al. 2000; Patrylo and Dudek 1998; Tauck and Nadler 1985). Recurrent mossy fiber growth may therefore break down the dentate filter and facilitate seizure propagation.
It has been suggested that the recurrent mossy fiber pathway also contributes to seizure propagation in another way: namely by compromising GABA inhibition through the release of zinc. Within the mossy fiber bouton, zinc is sequestered in synaptic vesicles (Fredrickson and Danscher 1990), and can be released in a Ca2+-dependent manner by depolarizing stimuli (Aniksztejn et al. 1987; Assaf and Chung 1984; Budde et al. 1997; Howell et al. 1984). At mossy fiber synapses on CA3 pyramidal cells, stimulus-induced release of zinc inhibits the opening ofN-methyl-d-aspartate (NMDA) channels by glutamate (Vogt et al. 2000). GABAA receptors are also sensitive to zinc. Both pilocarpine-induced status epilepticus (Gibbs et al. 1997) and kindling (Buhl et al. 1996) markedly increase the sensitivity to zinc of GABAAreceptors expressed by dentate granule cells, possibly due to altered subunit expression (Brooks-Kayal et al. 1998). In the kindling model of epilepsy, zinc (200 μM) reportedly reduces the frequency, mean amplitude, rate of rise and decay time constant of miniature inhibitory postsynaptic currents (mIPSCs) (Buhl et al. 1996). None of these effects has been observed in granule cells from control rats. Interestingly, GABAA receptor currents in granule cells dissociated from hippocampi surgically resected for medically intractable complex partial seizures resemble those from pilocarpine-treated epileptic rats in their sensitivity to zinc (Shumate et al. 1998). Thus release of zinc from the recurrent mossy fiber pathway, if it overflows the synapse in sufficient concentration, might disinhibit the postsynaptic granule cell. The objective of the present study was to evaluate this hypothesis.
Induction of recurrent mossy fiber growth
Adult male Sprague-Dawley rats (175–225 g; Zivic-Miller Laboratories, Allison Park, PA) received a single injection of pilocarpine hydrochloride (330–360 mg/kg ip) 30 min after an injection of scopolamine methyl bromide and terbutaline hemisulfate (2 mg/kg ip, each). Only rats that developed status epilepticus after pilocarpine administration were used in this study. Status epilepticus was terminated 3.5 h after onset with a single injection of phenobarbital sodium (50 mg/kg ip). Electrophysiological studies were performed 10–30 wk after pilocarpine administration.
Preparation of hippocampal slices
The rat was decapitated under ether anesthesia, the brain was removed and quartered, and 400-μm-thick transverse slices were cut through the caudal third of the hippocampal formation with a vibratome. Slices used for electrophysiological recording corresponded to horizontal plates 98–100 of Paxinos and Watson (1986). Additional slices reserved for Timm histochemistry were taken from a level of the hippocampal formation immediately rostral, corresponding to plates 101–103. For electrophysiological studies, slices were transferred to a beaker of artificial cerebrospinal fluid [standard ACSF, which contained (in mM) 122 NaCl, 25 NaHCO3, 3.1 KCl, 1.8 CaCl2, 1.2 MgSO4, 0.4 KH2PO4, and 10d-glucose, pH 7.4] and oxygenated at room temperature for ≥1.5 h with 95% O2-5% CO2. In some experiments, 200 nM ZnCl2 was added to the medium.
Effect of mossy fiber stimulation on GABAAreceptor-mediated postsynaptic currents
A slice was transferred to a glass-bottom Plexiglas submersion-type recording chamber mounted on the stage of a Nikon Optiphot-2 upright microscope (Nikon, Melville, NY) connected to a Noran Odyssey confocal imaging system (Noran Instruments, Middleton, WI). The chamber was filled with ACSF that was recirculated at a rate of 4 ml/min at room temperature (22–24°C). The total volume of superfusion medium was 10 ml. In most experiments, the ACSF used for superfusion was modified as follows. MgSO4 was replaced with MgCl2, KH2PO4 was omitted and the KCl concentration was increased to 3.5 mM (PO4/SO4-free ACSF). Patch electrodes were pulled from borosilicate glass (1.5 mm OD, 1.1 mm ID, Sutter Instruments, Novato, CA) and had a tip resistance of 5–7 MΩ. The tip of the electrode was filled by vacuum with a solution that contained (in mM) 140 cesium gluconate, 15 HEPES, 3.1 MgCl2, 1 CaCl2, and 11 EGTA, pH 7.2 and 276 mosm. The electrode was then backfilled with internal solution that contained (in mM) 120 cesium gluconate, 10 HEPES, 2 MgATP, 1 EGTA, 5 creatine phosphate, 10 N-ethyl lidocaine (QX-314 ) chloride, and 0.1 Alexa Fluor 488 hydrazide plus 20 units/ml creatine phosphokinase, pH 7.4 and 276 mosm.
Whole cell patch-clamp recordings were made from granule cells located in the infrapyramidal blade of the dentate gyrus because Timm histochemistry indicated that recurrent mossy fiber growth is often denser there than in the suprapyramidal blade (Okazaki et al. 1999). Gigaohm seals were formed by the “blind” approach (Blanton et al. 1989) on granule cell bodies located at least 30 μm below the upper surface of the slice. Whole cell access was obtained in current-clamp mode; only cells withV m > −70 mV on break-in (after correction for a 10-mV liquid junction potential) were accepted for study. Granule cell identity was confirmed by visualizing intracellular Alexa Fluor 488 (excitation: 488-nm, 515-nm barrier filter) and observation of strong spike-frequency adaptation during a suprathreshold depolarization.
To evoke the release of zinc from recurrent mossy fiber boutons, a bipolar stimulating electrode (25-μm-diam nichrome wires insulated to the tips with polymerized polyvinyl resin, tip separation of 0.3 mm) was placed in stratum lucidum of area CA3b ∼100 μm from the opening of the dentate hilus. Extracellular field recordings were used to optimize the position of the stimulating electrode, as previously described (Okazaki et al. 1999). Constant-current rectangular stimulus pulses of 100-μs duration were delivered with a Grass (W. Warrick, RI) stimulator and stimulus isolator every 10 s. Stimulus current (300–500 μA) was adjusted to evoke a just-maximal antidromic population spike.
Patch-clamp recordings were made with an Axon Instruments (Foster City, CA) Axopatch 1D amplifier beginning ∼20 min after achieving whole cell access. Series resistances ranged from 6 to 22 mΩ and were compensated ∼50%. Signals were filtered at 2 kHz, sampled at 20 kHz during the response to uncaged GABA and at 0.5 kHz during the rest of the recording, then stored to disk with use of a TL 1–125 digitizing board and PClamp6 (Axon Instruments). The excitatory postsynaptic current (EPSC) evoked by antidromic stimulation of the mossy fibers was recorded at a holding potential of −80 mV and was defined as the difference in the inward current before and after superfusion with 5 μM 2,3-dihydroxy-6-nitro-7-sulfamyl-benzo(F)quinoxaline (NBQX) and 50 μM d-2-amino-5-phosphonopentanoate (d-AP5). Granule cells in which mossy fiber stimulation failed to evoke an EPSC were not studied further.
An 80-mW Coherent Enterprise 653 argon ion ultraviolet (UV) laser (Coherent Laser Group, Santa Clara, CA) was used to release GABA from a caged precursor. The laser was coupled to the epifluorescence input of the microscope by a fiber optic cable, and the output passed through an Olympus (Melville, NY) water-immersion, UV-corrected ×40 objective (NA, 0.7; working distance, 3.2 mm). The effective diameter of the laser beam within the focal plane was 5.3 μm (Molnár and Nadler 1999). Shutter opening was controlled by PClamp6.
γ-Aminobutyrate, α-carboxy-2-nitrobenzyl ester (cGABA, 200 μM) was added to the superfusion medium and the laser beam was focused either on the soma of the recorded cell or on the apical dendrite 50, 75, or 100 μm from the soma. Then 4-ms pulses of UV light were applied at 10-s intervals to release GABA from the caged presursor. The laser power (20–50 mW at the source) was adjusted to evoke an outward current of 100–200 pA recorded at a holding potential of 0 mV. High-frequency stimulation (10 or 100 Hz for 1 s) was applied to the mossy fiber pathway just preceding every other uncaging event. Shutter opening was timed to occur just after the next to last stimulus pulse in the train. In some experiments, the mossy fiber pathway was stimulated at a frequency of 10 Hz for 5 min, and pulses of UV light were delivered every 10 s. GABAAreceptor-mediated outward currents recorded in the presence and absence of mossy fiber stimulation (n = 5 light flashes in each group) were averaged. Their peak amplitudes were measured with reference to the leak current just before shutter opening, and the effect of high-frequency stimulation was expressed as the percentage change in peak amplitude compared with the peak amplitude in the absence of mossy fiber stimulation. GABABreceptor-mediated currents did not contaminate our recordings because these currents were blocked by the use of a cesium-based internal solution that contained QX-314 but not GTP.
Some experiments were performed under conditions intended to enhance the release of zinc (and glutamate) from mossy fiber boutons. MgCl2 was omitted from the PO4/SO4-free ACSF, the KCl concentration was increased to 6 mM, and the temperature of the extracellular solution within the recording chamber was raised to 33°C. Glutamate antagonists were not used. To prevent interference from currents evoked by synaptically released glutamate, currents evoked by uncaged GABA were recorded atE EPSC. The value ofE EPSC was determined at the beginning of each experiment. The 6 mM K+/nominally Mg2+-free solution was superfused for ≥20 min before experimentation began. Laser-evoked photolysis of cGABA was carried out in the absence and presence of 10-Hz stimulus trains, as described in the preceding text.
Effect of zinc on muscimol-evoked GABAAreceptor-mediated currents
A glass micropipette (tip diameter ∼5–10 μm) that contained 400 μM muscimol and 100 μM fluorescein-dextran dissolved in standard ACSF was placed above the hippocampal slice over the inner third of the dentate molecular layer. Muscimol-containing solution was ejected from the micropipette with a Picospritzer (6–10 psi; General Valve, Fairfield, NJ) such that the solution contacted the surface of the slice just above the apical dendrite of the recorded cell. Muscimol was applied for 300–500 ms every 2 min.
Effect of zinc on mIPSCs
These experiments utilized a modified internal solution; cesium gluconate was replaced with CsCl and QX-314 was omitted. The use of CsCl-filled electrodes shifted E Cl to ∼0 mV. Thus GABAA receptor-mediated currents were inwardly directed. The superfusion medium contained 5 μM NBQX, 50 μM d-AP5 and 1 μM tetrodotoxin. mIPSCs were recorded at a holding potential of −70 mV during a 150-s epoch. The liquid junction potential was taken as 0 mV with a CsCl-based internal solution. Spontaneous events were recorded with PClamp6 and analyzed with functions incorporated in PClamp6 and Mini Analysis (Jaejin Software, Leonia, NJ). The threshold for detection of an mIPSC was 6 pA.
Effect of zinc on NMDA receptor-mediated EPSCs
A bipolar stimulating electrode was inserted into the perforant path where it crosses the subiculum. The NMDA receptor-mediated component of the perforant path EPSC was isolated pharmacologically by adding 5 μM NBQX and 30 μM bicuculline to the superfusion medium. Rectangular current pulses (100 μs duration) were applied every 30 s. The stimulus current was adjusted to evoke a 150- to 200-pA inward synaptic current recorded at a holding potential of −20 mV.
Adjacent slices from the same hippocampi were always processed for histochemical detection of heavy metal. In six experiments, the slice that had been used for electrophysiological recording was also studied in this way. Slices were immersed in 0.1% (wt/vol) Na2S, 0.1 M sodium phosphate buffer, pH 7.3, for 1.5 h followed by fixation in phosphate-buffered 10% formalin at 4°C for 1–2 days. They were then embedded in albumin-gelatin, and 30-μm-thick sections were prepared with a Vibratome. Slide-mounted sections were processed as described by Danscher (1981)and lightly counterstained with cresyl violet.
cGABA, Alexa Fluor 488 hydrazide and fluorescein dextran were purchased from Molecular Probes (Eugene, OR);N,N,N′,N′-tetrakis(2-pyridyl-methyl)ethylenediamine (TPEN), d-gluconic acid lactone, HEPES, EGTA, creatine phosphate, creatine phosphokinase, phenobarbital sodium, pilocarpine hydrochloride, (−)scopolamine methyl bromide, and terbutaline hemisulfate from Sigma Chemical (St. Louis, MO); d-AP5 from Tocris Cookson (Bristol, UK); bicuculline methiodide from Research Biochemicals (Natick, MA); and cesium hydroxide (99.9%; 50 wt%) from Aldrich (Milwaukee, WI). QX-314 chloride was obtained from Astra USA (Westborough, MA) and Alomone Labs (Jerusalem, Israel). NBQX was a gift from Novo Nordisk (Måløv, Denmark).
Effect of zinc on responses mediated by GABAA and NMDA receptors in standard and PO4/SO4-free ACSF
The purpose of these experiments was to confirm that zinc depressed GABAA receptor function in dentate granule cells from our pilocarpine-treated rats in accordance with previous reports (Brooks-Kayal et al. 1998; Buhl et al. 1996; Gibbs et al. 1997; Shumate et al. 1998). During superfusion with standard ACSF, we were unable to demonstrate any effect of 200 μM zinc on muscimol-evoked currents (Fig. 1 C) or mIPSCs (Table 1). Furthermore, zinc depressed the NMDA receptor-mediated component of the perforant path EPSC to only a minor degree (Fig. 1 D). Robust effects of zinc appeared when the polyvalent anions phosphate and sulfate were removed from the superfusion medium (Buhl et al. 1996). During superfusion with PO4/SO4-free ACSF, 200 μM zinc depressed muscimol-evoked currents by 23 ± 7% (mean ± SD, n = 6; Fig. 1, A and C). It reduced the frequency, mean amplitude, and charge transfer of mIPSCs without altering response kinetics (Table 1, Fig.2). Finally, it reduced the peak amplitude of the NMDA receptor-mediated component of the perforant path EPSC by 81 ± 6% (Fig. 1, B and D).
High-frequency stimulation of the recurrent mossy fiber pathway did not alter responses to uncaged GABA
All granule cells included in this study responded to mossy fiber stimulation with an EPSC. The mean peak amplitude of the EPSC recorded at a holding potential of −80 mV was 205 pA, and individual responses ranged in amplitude from 10 to 1,500 pA. cGABA did not affect the size of the EPSC (Molnár and Nadler 2000). However, mossy fiber stimulation evoked very small IPSCs under these conditions (Fig. 3 B) (Molnár and Nadler 2000).
Previous reports suggested that high-frequency stimulus trains more effectively release zinc from the mossy fiber pathway than low-frequency stimulation (Aniksztejn et al. 1987;Assaf and Chung 1984). We therefore tested the effect of high-frequency mossy fiber stimulation on currents evoked by GABA. GABA was released onto the granule cell dendrite adjacent to recurrent mossy fiber synapses by laser-evoked photolysis of cGABA. We varied the stimulus frequency (10 or 100 Hz for 1 s). Experiments were carried out during superfusion with standard ACSF and with PO4/SO4-free ACSF. In some experiments, 200 nM ZnCl2, the approximate concentration present in human cerebrospinal fluid (Palm et al. 1983), was added to the superfusion medium to replenish any zinc that might be lost from the tissue. The laser beam was focused on the apical dendrite of the recorded cell 50, 75, or 100 μm from the soma, within the segment of dendrite innervated by recurrent mossy fibers (Molnár and Nadler 1999). For comparison, some measurements were made with the laser beam focused on the soma of the recorded cell, where the density of mossy fiber boutons is much lower. Under none of these conditions did mossy fiber stimulation significantly alter the response to uncaged GABA (Fig. 3, Band C; Table 2). There was, on average, a small, but not statistically significant (P> 0.05; paired t-test), reduction of the GABAA receptor-mediated current whenever the mossy fibers were stimulated at a frequency of 100 Hz (Fig.3 C). This trend in the data did not depend on mossy fiber zinc. The reduction of the GABAAreceptor-mediated current was comparable in the presence (13.5 ± 22.6%, n = 7) and absence (13.8 ± 24.8%,n = 7) of 100 μM TPEN, a membrane-permeant chelator of heavy metals with a high affinity for zinc (Arslan et al. 1985). Moreover, the same reduction of GABAA receptor-mediated current could be observed whether the laser beam was focused on the proximal portion of the dendrite or on the soma of the recorded cell.
Mossy fiber stimulation at a frequency of 100 Hz, but not at a frequency of 10 Hz, evoked a steady outward current in ∼50% of the granule cells tested (Fig. 3 C). The amplitude of this current varied considerably; in some experiments, it was comparable to the current evoked by uncaged GABA. This current was unrelated to mossy fiber zinc because addition of 100 μM TPEN to the superfusion medium did not affect its amplitude. Values were 79.9 ± 20.0 pA in the absence of TPEN and 78.4 ± 30.9 pA in the presence of TPEN (means ± SD, n = 7). There was an inverse correlation between the amplitude of the outward current evoked by a 100-Hz stimulus train and the peak amplitude of the current evoked by uncaged GABA (P < 0.001 by linear regression,n = 31). This relationship did not depend on mossy fiber zinc because it was unchanged by 100 μM TPEN.
In additional experiments, mossy fibers were stimulated at a frequency of 10 Hz for 5 min. In eight of these experiments, the slice was superfused with PO4/SO4-free ACSF without added zinc. About 1 min after the onset of stimulation, the amplitude of the GABAA receptor-mediated outward current declined slightly (7.1 ± 6.2%, n = 8) and remained depressed beyond the end of the stimulus train (Fig.3 D). This effect was observed equally well in the presence of 100 μM TPEN (7.2 ± 10.6%, n = 7). Similar results were obtained from six experiments with PO4/SO4-free ACSF and 200 nM ZnCl2, four experiments with standard ACSF, and three experiments with standard ACSF and 200 nM ZnCl2.
In view of these negative findings, we changed the experimental conditions in ways that were expected to enhance the release of zinc. Mg2+ was omitted from the PO4/SO4-free ACSF. [K+]o was raised to 6 mM so that mossy fiber stimulation would evoke epileptiform granule cell discharge (Fig. 4 A) (Hardison et al. 2000; Patrylo and Dudek 1998). Finally the temperature of the slice chamber was increased to 33°C. Under these conditions, a single just-maximal stimulus applied to the mossy fiber pathway evoked epileptiform granule cell discharge in four of the six slices tested. In five additional slices switching from PO4/SO4-free ACSF to nominally Mg2+-free medium with 6 mM K+ markedly increased both the peak amplitude and total charge transfer of the recurrent mossy fiber EPSC (Fig.4 B). Peak amplitude increased from 122 ± 69 to 413 ± 383 pA, and total charge transfer increased from 2.6 ± 1.6 to 18.3 ± 11.2 nC. These slices were used to test the effect of high-frequency mossy fiber stimulation on the response to uncaged GABA. Stimulation at 10 Hz for 1 s did not significantly alter this response, whether the laser beam was focused on the apical dendrite 75 μm from the soma or on the soma itself (Fig.4 C; Table 2).
cGABA did not alter the depression of GABAAreceptor-mediated currents by zinc
cGABA, at the concentration used in the present study (200 μM), blocks inhibitory synaptic transmission onto dentate granule cells (Molnár and Nadler 2000). Thus inhibitory synaptic currents, either spontaneous or evoked, were very small and should have minimally affected responses to applied muscimol or uncaged GABA. Because we do not understand fully the mechanism by which cGABA interferes with GABA transmission, we had to consider the possibility that cGABA modified the response of GABAAreceptors to zinc. No such action of cGABA was found, however. During superfusion with PO4/SO4-free ACSF, exposure to 200 μM zinc reduced muscimol-evoked currents by 23 ± 7 and 28 ± 12% (n = 6) in the absence and presence, respectively, of 200 μM cGABA.
Maintenance of mossy fiber zinc during stimulus trains
Seizure activity markedly reduces the zinc content of the mossy fiber pathway (Fredrickson et al. 1988; Sloviter 1985). Thus the high-frequency stimulus trains employed in the present study may have failed to alter responses to uncaged GABA because they depleted the mossy fibers of zinc before the uncaging event. We used Timm histochemistry to evaluate this possibility. When six slices in which the mossy fibers had been electrically stimulated at 10 Hz for 5 min were compared with slices processed for Timm histochemistry immediately after preparation, no difference in the extent or density of mossy fiber-like supragranular staining was evident (Fig. 5). Therefore the stimulus trains employed in this study did not deplete the mossy fiber boutons of zinc.
In all slices, dense mossy fiber-like Timm staining was observed in the inner third of the dentate molecular layer. This finding confirmed the presence of recurrent mossy fiber boutons at the sites of uncaging.
In animal models of epilepsy (Buhl et al. 1996;Gibbs et al. 1997) and possibly also in humans with temporal lobe epilepsy (Shumate et al. 1998), exogenously applied zinc much more potently inhibits the activation of GABAA receptors on dentate granule cells than it does normally. This study tested the hypothesis that zinc released from the recurrent mossy fiber pathway in response to electrical stimulation reaches an extracellular concentration sufficient to produce such inhibition. Our results suggest that it did not.
Effects of exogenously applied zinc on responses evoked by activation of GABAA and NMDA receptors
Our results confirmed earlier findings that GABAA receptors on dentate granule cells in rat models of epilepsy are sensitive to exogenously applied zinc. The inhibitory effects of zinc depended entirely on the removal of polyvalent anions from the superfusion medium, presumably because these anions bind zinc and render it inactive. The sensitivity of GABAA receptors to zinc also depends on subunit composition. Recombinant GABAA receptors that contain α4, α5, and α6 subunits are more sensitive to zinc than those that contain α1 subunits (Fisher and Macdonald 1998; Knoflach et al. 1996;Saxena and Macdonald 1996; White and Gurley 1995). Furthermore, the presence of a γ or ε subunit reduces and the presence of a δ subunit enhances zinc sensitivity (Draguhn et al. 1990; Saxena and Macdonald 1994). In dentate granule cells of the epileptic brain, enhanced zinc sensitivity may be explained by a greater expression of α4 subunits coupled with reduced expression of α1 subunits (Brooks-Kayal et al. 1998). However, 200 μM zinc only modestly depressed the response to exogenously applied agonist in our study; the muscimol-evoked outward current declined by an average of only 23%. This figure contrasts with the report of Gibbs et al. (1997) in which 100 μM zinc reduced GABA-evoked currents by 76% and 300 μM zinc reduced them by 91% in isolated dentate granule cells from pilocarpine-treated rats. Although further studies are needed to explain this discrepancy, one possibility is that GABAA receptors differ in zinc sensitivity according to their location on the granule cell. In previous studies, the entire isolated granule cell was exposed to GABA, whereas in the present study muscimol was applied to the inner third of the granule cell dendrite. Thus the GABA current in previous studies was generated predominantly by the opening of GABAA channels on the soma, whereas the muscimol current in the present study was generated predominantly by the opening of GABAAchannels on the proximal portion of the dendrite. Some results suggest that different GABAA receptor subtypes (possibly with different affinities for zinc) can be expressed to different degrees on different parts of the same neuron (Brickley et al. 1999). It is also possible, although less likely, that the granule cell isolation procedure altered the sensitivity of GABAA receptors to zinc.
Zinc (200 μM) also reduced the frequency, mean amplitude, and charge transfer of mIPSCs. These findings could be explained by a block of GABAA receptor activation, with the amplitude of the postsynaptic response sometimes being reduced to the extent that it could not be distinguished from background noise. Thus exogenously applied zinc gained access to synaptic GABAAreceptors, as well as to receptors (that could have been both synaptic and extrasynaptic) responsive to exogenous agonist.
mIPSC amplitude and frequency were less affected by zinc in the present study than in the study of Buhl et al. (1996) on dentate granule cells from kindled rats. Furthermore we did not reproduce their finding of a reduced rate of rise and faster decay time constant.Buhl et al. (1996) found no such effects of zinc on mIPSCs from control granule cells. It seems likely therefore that the changes in mIPSC amplitude and frequency that we observed, like the changes reported in the kindling model, reflected a modification of GABA synaptic function related to epileptogenesis. Evidently the nature of this modification differs somewhat among different epilepsy models.
Electrically evoked release of zinc did not alter the response to focally applied GABA
We used several approaches to maximize the chance of observing an effect of mossy fiber zinc on GABAA receptor activation in dentate granule cells. First, we used stimulus trains similar to those other investigators have shown to optimize the release of zinc (Aniksztejn et al. 1987). Second, we applied stimulus trains under conditions (nominally Mg2+-free medium, 6 mM [K+]o, 33°C) that clearly enhanced the release of glutamate and were expected also to enhance the release of zinc. Third, we limited the region of GABA exposure to just that portion of the dendritic tree contacted by recurrent mossy fiber boutons. Timm histochemistry confirmed that photolysis of cGABA took place in the region of highest recurrent mossy fiber density (50–100 μm from the soma of the recorded cell). Fluorescence imaging of zinc withN-(6-methoxy-8-quinolyl)-p-carboxybenzoylsulfonamide (TFLZn) (Budde et al. 1997) confirmed the presence of mossy fiber boutons in the supragranular zone (not shown). Fourth, we attempted to replenish any zinc lost from the mossy fiber boutons as a result of electrical stimulation. For this purpose, experiments were conducted in the presence of 200 nM zinc, a concentration close to that reported in human cerebrospinal fluid (Palm et al. 1983). All of these approaches proved fruitless. No effect of mossy fiber stimulation was ever observed on the granule cell's response to uncaged GABA, at least none that could be attributed to the release of zinc.
The validity of our experimental approach depends on the ability of electrical stimulation to release zinc from mossy fiber boutons in the hippocampal slice. Recently, Vogt et al. (2000) reported that the release of endogenous zinc by single electrical stimuli depresses the NMDA receptor-mediated component of the mossy fiber synaptic response recorded from CA3 pyramidal cells. The vesicular pool of zinc visualized by Timm histochemistry appears not to be the only source of releasable zinc (Lee et al. 2000). Thus it is not clear that mossy fiber stimulation releases zinc through exocytosis. Nevertheless, zinc released by some process does act on postsynaptic NMDA receptors. We have replicated the most pertinent results of Vogt et al. (2000) in studies of mossy fiber-granule cell synapses. That is, calcium disodium EDTA (CaEDTA), a high-affinity, membrane-impermeant zinc chelator, significantly increased the size of the NMDA component of the recurrent mossy fiber EPSC, this effect was observed only at negative holding potentials, and CaEDTA did not significantly change the size of the AMPA/kainate receptor-mediated component of the recurrent mossy fiber EPSC or the NMDA receptor-mediated component of the perforant path EPSC (Molnár and Nadler, unpublished data). There was no drop-off in the effect of CaEDTA during the experiment, suggesting that electrical stimulation does not easily deplete the releasable pool of zinc (see also Budde et al. 1997). In addition, imaging studies provide evidence for zinc release from mossy fiber boutons during high-frequency stimulus trains (Budde et al. 1997; Li et al. 2000; Quinta-Ferreira et al. 2000). Thus we conclude that the inability of mossy fiber stimulation to diminish the response to uncaged GABA did not result from failure of the pathway to release zinc.
The hypothesis that release of zinc from recurrent mossy fiber boutons inhibits GABAA receptor function requires that zinc overflow the mossy fiber synapse in sufficient concentration to inhibit the activation of GABAA receptors located at synapses nearby. The report of Gibbs et al. (1997)suggests that a local zinc concentration ≤10 μM could inhibit receptor activation to a measurable degree. Because the concentration of zinc in the mossy fiber synaptic cleft has been suggested to reach 100–300 μM during tetanically evoked release (Fredrickson and Danscher 1990), it may seem reasonable that its concentration at nearby GABA synapses would approach 10 μM. This hypothesis does not consider at least two factors that impede the diffusion of zinc away from the synaptic cleft. The first is the presence of zinc transporters on the plasma membrane of the mossy fiber bouton (Howell et al. 1984). These transporters would be expected to minimize zinc overflow through binding and subsequent reuptake. The second is the presence of polyvalent anions, which bind zinc and render it incapable of acting on receptors. Removal of polyvalent anions from the superfusion medium clearly enhanced the effects of zinc in the present study. Indeed we could not detect any depression of GABAA receptor activation without modifying the superfusion medium in this way. Yet polyvalent anions are present in the extracellular fluid of brain and may well limit the effects of zinc, especially outside the mossy fiber synaptic cleft where the zinc concentration would be low. Negative surface charges may also impede the diffusion of zinc. These considerations, in conjunction with our experimental findings, suggest that zinc in the vicinity of granule cell GABAA receptors may not reach a concentration sufficient to diminish receptor activation significantly.
Two limitations of our experimental approach should be noted. First, the mossy fiber stimulation we used did not activate the entire recurrent pathway but only some fraction of it; some of the mossy fibers in stratum lucidum undoubtedly coursed out of the plane of the slice before reaching the dentate molecular layer. Thus our experiments fell short of modeling in vivo conditions in which the entire granule cell population is driven to fire synchronously. More zinc is likely to be released from mossy fiber boutons in this condition than in response to the stimuli used in the present study. Synchronous high-frequency granule cell discharge is a rare event, however; granule cells normally fire action potentials asynchronously at low rates (Jung and McNaughton 1993). In the epileptic brain, expansion of the recurrent mossy fiber pathway (Cronin et al. 1992;Hardison et al. 2000; Patrylo and Dudek 1998; Tauck and Nadler 1985) and slightly elevated [K+]o(Hardison et al. 2000) facilitate granule cell synchrony. It seems unlikely that zinc-induced block of GABA inhibition could play a significant role in this process, however, because synchronous high-frequency granule cell discharge would probably have to be present already in order for zinc to reach a concentration at GABA synapses that is sufficient to produce significant disinhibition. Second, laser-evoked photolysis of GABA exposes both synaptic and extrasynaptic GABAA receptors to agonist. We cannot be certain how much each type contributed to the GABA current we recorded. It is possible that a significant effect of zinc on synaptic GABAA receptors was masked by a lack of effect on extrasynaptic receptors. Unfortunately, there is no simple way at present to study the effect of mossy fiber zinc on GABA synapses immediately adjacent to the sites of zinc release.
We thank K. Gorham for secretarial assistance.
This research was supported by National Institute of Neurological Disorders and Stroke Grants NS-17771 and NS-38108.
Address for reprint requests: J. V. Nadler, Dept. of Pharmacology and Cancer Biology, Duke University Medical Center, Durham, NC 27710 (E-mail:).
- Copyright © 2001 The American Physiological Society