Effects of glutamate receptor agonists and antagonists on Ca2+ uptake in rat hippocampal slices lesioned by glucose deprivation or by kainate

Pergamon

PII:

Neuroscience Vol. 77, No. 1, pp. 97–109, 1997 Copyright ? 1997 IBRO. Published by Elsevier Science Ltd Printed in Great Britain 0306–4522/97 $17.00+0.00 S0306-4522(96)00426-5

EFFECTS OF GLUTAMATE RECEPTOR AGONISTS AND ANTAGONISTS ON Ca2+ UPTAKE IN RAT HIPPOCAMPAL SLICES LESIONED BY GLUCOSE DEPRIVATION OR BY KAINATE K. ALICI*, T. GLOVELI, D. SCHMITZ and U. HEINEMANN Department of Neurophysiology, Institute of Physiology, Charite´, Medical School of the Humboldt University Berlin, Tucholskystr. 2, 10117 Berlin, Germany Abstract––The functional relevance of presynaptic glutamate receptors in controlling presynaptic Ca2+ influx and thereby transmitter release is unknown. To test if presynaptic Ca2+ entry in the hippocampus is controlled by glutamate autoreceptors, we created a hippocampal slice preparation for investigation of presynaptic Ca2+ signals with Ca2+-sensitive microelectrodes after lesioning of neurons by glucose deprivation or kainate. Stratum radiatum and alveus stimulation-induced postsynaptic field potential components were irreversibly abolished in areas CA1 and CA3 of lesioned slices, whereas stratum radiatum stimulation still evoked afferent volleys. Repetitive stimulation of the stratum radiatum still induced decreases in extracellular Ca2+ concentration. Repetitive stimulation of the alveus no longer induced decreases in extracellular Ca2+ concentration, suggesting complete damage of pyramidal cells. The stratum radiatum stimulation-induced decreases in extracellular Ca2+ concentration in lesioned slices were comparable to those elicited during application of the glutamate antagonists 6-cyano-7nitroquinoxaline-2,3-dione and -2-amino-5-phosphonovalerate. In lesioned slices the stimulus-induced presynaptic Ca2+ influx was reversibly reduced by kainate, RS-á-amino-3-hydroxy-5-methyl-4-isoxazolepropionate (AMPA), N-methyl--aspartate and glutamate without effects on afferent volleys. The kainate and N-methyl--aspartate effects on presynaptic Ca2+ signals were partly sensitive to 2,3-dihydroxy-6-nitro-7-sulphamoyl-benzo(f)quinoxaline and -2amino-5-phosphonovalerate, respectively, while the AMPA effects were not significantly affected by 2,3-dihydroxy-6-nitro-7-sulphamoyl-benzo(f)quinoxaline, suggesting involvement of a novel glutamate receptor subtype. The involvement of a novel glutamate receptor subtype was supported by our findings that ionotropic glutamate receptor agonists also reduce presynaptic Ca2+ influx under conditions of blocked synaptic transmission by 6-cyano-7-nitroquinoxaline-2,3-dione and -2-amino-5phosphonovalerate. 1-Aminocyclopentane-trans-1,3-dicarboxylic acid had no significant effect on presynaptic Ca2+ entry. Also, the presynaptic Ca2+ influx was not influenced by the glutamate receptor antagonists 6-cyano-7-nitroquinoxaline-2,3-dione, 2,3-dihydroxy-6-nitro-7-sulphamoylbenzo(f)quinoxaline and -2-amino-5-phosphonovalerate when applied alone. Low kainate concentrations (5 µM) reduced presynaptic Ca2+ signals in area CA3 but not in area CA1, demonstrating the higher affinity of presynaptic kainate receptors on mossy fibre terminals. Copyright ? 1997 IBRO. Published by Elsevier Science Ltd. Key words: presynaptic Ca2+ influx, glutamate autoreceptors, glucose deprivation, kainate, hippocampal slice, rat.

It has been shown previously that neurotoxic treatments such as glucose deprivation3,10,55 or application of high concentrations of different glutamate receptor agonists27,39,40,45,53 lead to acute neuronal cell death, sparing afferent fibres, presynaptic terminals and presumably glial cells,44 in rat hippocampal

slices. In such acutely lesioned slices, stimulation of Schaffer collateral–commissural fibres in the stratum radiatum (SR) still elicits afferent volleys and ionic changes, including small decreases in extracellular Ca2+ concentration ((Ca2+)o), presumably due to influx into presynaptic terminals.1 These decreases in (Ca2+)o are similar to those observed after Schaffer collateral–commissural fibre stimulation under conditions of blocked synaptic transmission due to simultaneous application of 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) and -2-amino-5-phosphonovalerate (2-APV), receptor antagonists for RS-á-amino-3-hydroxy-5-methyl-4isoxazolepropionate (AMPA)/kainate and N-methyl-aspartate (NMDA) receptors for glutamate, respectively.1

*To whom correspondence should be addressed. Abbreviations: ACSF, artificial cerebrospinal fluid; AMPA, RS-á-amino-3-hydroxy-5-methyl-4-isoxazolepropionate; 2-APV, -2-amino-5-phosphonovalerate; (Ca2+)o, extracellular Ca2+ concentration; CNQX, 6-cyano-7-nitroquinoxaline-2,3-dione; NBQX, 2,3-dihydroxy-6-nitro7-sulphamoyl-benzo(f)quinoxaline; NMDA, N-methyl-aspartate; SP, stratum pyramidale; SR, stratum radiatum; trans-ACPD, 1-aminocyclopentane-trans-1,3dicarboxylic acid. 97

98

K. Alici et al.

We have shown previously that such decreases in (Ca2+)o are sensitive to GABA,21,28 baclofen21,29 and adenosine52 in area CA1 of the hippocampus. In the dentate gyrus, decreases in (Ca2+)o induced by perforant path stimulation are, in contrast, insensitive to baclofen9 and norepinephrine.57 Moreover, these decreases are augmented by glucose deprivation1 and blocked by oxygen deprivation67 in area CA1. It is well known that glutamate receptors are also located on presynaptic endings.33,41,43,50,56 The function of these presynaptic receptors is presently unknown. However, in analogy to other transmitter systems such as the monoaminergic systems,9 it might be speculated that they serve a function in controlling glutamate release. Such a system serving as a regulatory site for presynaptic transmitter release might be an interesting target for the development of neuroprotective and anticonvulsant drugs. Here, we report that glutamate receptor agonists reduce presynaptic Ca2+ entry in slices acutely lesioned either by glucose deprivation or by kainate, as well as under conditions of blocked synaptic transmission. EXPERIMENTAL PROCEDURES

The experiments were performed on more than 130 rat hippocampal slices. The slices were obtained from Wistar rats of both sexes (bred in-house; 180–200 g), decapitated under ether anaesthesia. Transverse hippocampal slices of nominally 350 µm were prepared as described previously,34 transferred into an interface chamber (35&1)C) and continuously perfused with oxygenated (95% O2/5% CO2), prewarmed artificial cerebrospinal fluid (ACSF) containing (in mM): NaCl 124, KCl 3, CaCl2 1.6, MgCl2 1.8, NaHCO3 26, NaH2PO4 1.25, glucose 10, pH 7.4. After 1 h of recovery, control recordings were performed to document the normal shape of the SR and alveus-induced field potential. Subsequently, the ACSF was replaced by a glucose-free ACSF, which was applied for 60–90 min. Alvear stimulation caused no signals except for stimulus artefacts in such preparations. Other slices were treated with 5–10 mM kainate in order to achieve the same goal, i.e. disappearance of potential components attributable to postsynaptic ion fluxes. For stimulating afferent or efferent fibres of hippocampal pyramidal cells of area CA1 or CA3, bipolar platinum wire

stimulating electrodes were inserted into the SR, the alveus, the hilus or the fimbria. Stimulus trains consisted of 0.1-ms, 5–10-V square pulses applied at 20 Hz for 10 s. Ca2+sensitive electrodes were manufactured and tested as described previously,22,32 and inserted into the stratum pyramidale (SP) and/or the SR of area CA1 or CA3 at a depth of 100–150 µm below the cut surface of the slice. Materials Drugs were bath applied by dissolving sodium salts of CNQX (30 µM), 2,3-dihydroxy-6-nitro-7-sulphamoylbenzo(f)quinoxaline (NBQX; 10 µM), 2-APV (30 µM), kainate (5 µM–10 mM), AMPA (100 µM), 1-aminocyclopentane-trans-1,3-dicarboxylic acid (trans-ACPD; 50 µM), NMDA (50–500 µM), -glutamic acid (glutamate; 10 mM). Glutamate, AMPA, NMDA and trans-APCD were purchased from Sigma (Deisenhofen, Germany), kainate, CNQX and 2-APV were obtained from Tocris (Bristol, U.K.). NBQX was a gift from Dr T. Honore´ at Novo Nordisk (Denmark). In other experiments the same drugs dissolved in ACSF were applied in droplets from a microinjection syringe (26-gauge), which permitted application of volumes of 10 µl. The droplet was applied in the vicinity of the recording electrodes, but rapidly spread over a large area of the slice as judged by addition of a food dye (E102, E122, E124 or E142). The droplets contained 100 mM glutamate, 1 mM kainate, 1 mM AMPA, 1 mM NMDA, 1 mM 2-APV or 1 mM NBQX dissolved in normal ACSF. As a control, drug-free droplets of ACSF were applied. This never resulted in any significant effect on the stimulus-induced changes in (Ca2+)o or on stimulus-induced field potentials. 2-APV and NBQX were either applied alone or in combination with the different glutamate agonists. The choice of the drug concentrations in the droplet were based on measurements with droplet applications of different Ca2+ concentrations in ACSF. The measurements of the changes of (Ca2+)o were performed with Ca2+-sensitive electrodes at a depth of 100–150 µM, where stimulus-induced changes in (Ca2+)o and field potentials were also measured. After the droplet application of Ca2+ at concentrations of 2.5 (n=6), 5 (n=4), 10 (n=6), 25 (n=4) and 50 mM (n=6), 5.6%, 6.2%, 6.0%, 8.6% and 9.4% of the applied Ca2+ concentration, respectively, were detected by the Ca2+-sensitive microelectrode. Hence, less than 10% of the Ca2+ applied to the surface of the slice appeared near the recording site. The peak of (Ca2+)o was reached after 1–2 min and the (Ca2+)o recovered to baseline in about 15 min. Similar observations were made with droplet application of tetramethylammonium, which was measured with nominally K+-sensitive

Fig. 1. Effects of CNQX/2-APV, glucose deprivation and high kainate application on SR stimulation (SR-STIM)-induced changes in (Ca2+)o and transient field potentials in area CA1. (Aa–c) Sample recordings of changes in (Ca2+)o and transient field potentials in the SR elicited by SR stimulation before and during application of 30 µM CNQX/30 µM 2-APV (Aa), before and after 60 min glucose deprivation (Ab), and before and after application of 8 mM kainate for 30 min (Ac). Note loss of stimulus-induced postsynaptic responses during application of CNQX and 2-APV, after glucose deprivation and after high kainate application, while sparing afferent volleys and the presumed presynaptic Ca2+ signal. Stimulus parameters for (Ca2+)o recordings: 20 Hz, 0.1 ms, 7 V, 10 s. Stimulus parameters for transient field potentials: 7 V, 0.1 ms, 50 ms interval. Asterisks indicate single stimuli, horizontal bars a stimulus train. (B) Averages of SR stimulation-induced changes in (Ca2+)o (Ä(Ca2+)o) recorded in the SR before and in the presence of the glutamate receptor antagonists 30 µM CNQX/30 µM 2-APV (n=13; shaded columns), before and after 60 min glucose deprivation (n=26; open columns) and before and after 5–10 mM kainate application (n=14; filled columns). (C) Averages of SR stimulation-induced changes in (Ca2+)o (Ä(Ca2+)o) recorded in the SP before and in the presence of the glutamate receptor antagonists 30 µM CNQX/30 µM 2-APV (n=9; shaded columns), before and after 60 min glucose deprivation (n=26; open columns), and before and after 5–10 mM kainate application (n=14; filled columns). Note that after lesioning by glucose deprivation or high kainate application and also after blockade of synaptic transmission by CNQX/2APV, SR stimulation leads to greatest decreases in (Ca2+)o in the SR. Mean values and S.D. are shown.

Effects of glutamate receptor agonists on Ca2+ uptake

Fig. 1.

99

100

K. Alici et al.

electrodes (Heinemann U. and Dietzel I., unpublished observations). Consequently, we assume that drugs with little cellular uptake reach maximally 10% of the applied concentration in the recording position. All data were displayed on a chart writer. All transient field potentials and often also changes in (Ca2+)o and associated slow field potentials were stored on a computer disk for further evaluation. Changes in (Ca2+)o were evaluated using the formula: ÄCa2+=(Ca2+)B1((Ca2+)B10(ÄE/S)), with (Ca2+)B being the baseline calcium concentration, ÄE being the potential change measured with the Ca2+-selective electrode, S being the slope for the Ca2+-sensitive microelectrode determined with 10-fold changes in (Ca2+)o. The total Ca2+ concentration in the ACSF was 1.6 mM. The free Ca2+ concentration was reduced by bicarbonate and phosphate buffers by roughly 25% to the physiological baseline of 1.2 mM.20 All results are expressed as mean values&S.D. with n=number of measurements. Statistical significance was determined using Student’s t-test. Differences were accepted as statistically significant if P was less than 0.05.

RESULTS

Stimulus-induced changes in (Ca2+)o under blocked excitatory synaptic transmission in strata radiatum and pyramidale in hippocampal slices As well as under conditions of blocked synaptic transmission by low-Ca2+ ACSF,24,28,51 during blockade of excitatory synaptic transmission by CNQX/2-APV repetitive stimulation of the SR induced the greatest (Ca2+)o decreases in the SR, the main synaptic input zone of hippocampal pyramidal cells (Fig. 1B, C). CNQX (30 µM) and 2-APV (30 µM) bath applied for 30 min reduced SR stimulus-induced decreases in (Ca2+)o in both the SR and SP by 87&13% (n=13) and 90&10% (n=9), respectively (Fig. 1Aa, B, C). These effects were largely reversible. Changes in (Ca2+)o evoked by alvear stimulation are also reversibly reduced by CNQX and 2-APV. However, the effect is less drastic in the SR and SP (Fig. 2Aa, B, C).

Stimulus-induced changes in (Ca2+)o in lesioned hippocampal slices After 60 min glucose deprivation or after application of 5, 8 or 10 mM kainate for 30 min, stimulation of the SR still elicited afferent volleys, and during repetitive stimulation small decreases in (Ca2+)o followed by overshoots were noted (Fig. 1). These (Ca2+)o changes were similar in size to those under conditions when excitatory synaptic transmission was blocked by CNQX/2-APV. In contrast, stimulation of the alveus no longer evoked any signals in (Ca2+)o, in slow and transient field potentials (Fig. 2). During 60 min of glucose withdrawal from the ACSF, we observed a slow fall in baseline (Ca2+)o to 0.59&0.12 mM (n=25), but no significant change in the field potential baseline in area CA1. The (Ca2+)o baseline returned to normal levels within 1–2 h after return to glucose-containing ACSF. Application of 5, 8 or 10 mM kainate for 30 min also evoked a fall of baseline (Ca2+)o. In this case the decreases in (Ca2+)o were always associated with a negative shift in the field potential baseline. Kainate (5 mM) application resulted in a rapid drop of baseline (Ca2+)o to 0.33&0.3 mM, associated with a negative shift of the field potential baseline by 5.2&2.5 mV (n=6) in area CA1. Kainate (8 or 10 mM) did not result in larger negative shifts of the field potential baseline nor in larger drops of (Ca2+)o. After return to perfusion with normal medium, the field potential baseline and (Ca2+)o recovered to normal levels within about 1 h. Effects of various glutamate receptor agonists and antagonists on changes in (Ca2+)o induced by repetitive stimulation of the stratum radiatum in lesioned slices When glutamate, kainate, AMPA, NMDA or trans-ACPD were bath or droplet applied, this no longer resulted in any detectable decreases in (Ca2+)o nor in any slow field potential shifts. Figure 3 illustrates the reduction of the decreases in (Ca2+)o in slices pretreated with glucose-free ACSF in area

Fig. 2. Effects of CNQX/2-APV, glucose deprivation and high kainate application on alveus stimulation (ALV-STIM)-induced changes in (Ca2+)o and transient field potentials in area CA1. (Aa–c) Sample recordings of changes in (Ca2+)o and transient field potentials in the SR elicited by alveus stimulation before and during application of 30 µM CNQX/30 µM 2-APV (Aa), before and after 60 min glucose deprivation (Ab), and before and after application of 8 mM kainate (Ac). Note loss of stimulus-induced postsynaptic responses after glucose deprivation and after high kainate application. Stimulus parameters for (Ca2+)o recordings: 20 Hz, 0.1 ms, 7 V, 10 s. Stimulus parameters for transient field potentials: 7 V, 0.1 ms, 50 ms interval. Asterisks indicate single stimuli, horizontal bars a stimulus train. (B) Averages of alveus stimulation-induced changes in (Ca2+)o (Ä(Ca2+)o) recorded in the SR before and in the presence of the glutamate receptor antagonists 30 µM CNQX/30 µM 2-APV (n=13; shaded columns), before and after 60 min glucose deprivation (n=26; open columns), and before and after 5–10 mM kainate application (n=14; filled columns). (C) Averages of alveus stimulation-induced changes in (Ca2+)o (Ä(Ca2+)o) recorded in the SP before and in the presence of the glutamate receptor antagonists 30 µM CNQX/30 µM 2-APV (n=9; shaded columns), before and after 60 min glucose deprivation (n=26; open columns), and before and after 5–10 mM kainate application (n=14; filled columns). Note that after lesioning procedures alveus stimulation no longer induced decreases in (Ca2+)o or population spikes. Mean values and S.D. are shown.

Effects of glutamate receptor agonists on Ca2+ uptake

Fig. 2.

101

102

K. Alici et al.

Fig. 3. Effects of bath-applied kainate on presumed presynaptic Ca2+ entry in area CA1 in a slice lesioned by 60 min glucose deprivation. (Aa1, a2) SR stimulation (SR-STIM)-induced changes in (Ca2+)o and transient field potentials in normal ACSF. (Ba–c) After irreversible loss of postsynaptic signal components caused by glucose deprivation, 1 mM kainate depressed the SR stimulation-induced changes in (Ca2+)o significantly and reversibly. During application of kainate there was no change in baseline (Ca2+)o or blockade of afferent volleys. Asterisks indicate single stimuli and horizontal bars a stimulus train. Voltage calibration in Bc2 applies to Ba2 and Bb2.

Effects of glutamate receptor agonists on Ca2+ uptake

CA1. Kainate (1 mM) bath applied for 30 min depressed the changes in (Ca2+)o evoked by repetitive stimulation of the SR reversibly by 70&12% (n=11). Afferent volleys elicited by SR stimulation were not affected by application of kainate (Fig. 3B). Bath applied kainate (100 µM) reduced the SR repetitive stimulation-induced decreases in (Ca2+)o by 50&11% (n=7), whereas 10 µM bath applied kainate had no effect (1&2%; n=5). When 1 mM kainate was applied in a droplet, the SR stimulus-induced decreases in (Ca2+)o were reduced by 38&17% (n=16). Droplet application of kainate in kainate-lesioned slices resulted in similar reductions of the SR-induced Ca2+ signals. In order to test the possible involvement of NBQX-sensitive glutamate receptors, the NBQX was bath applied. NBQX (10 µM) bath applied for 30 min did not affect stimulus-induced Ca2+ signals. When droplets containing 1 mM kainate and 1 mM NBQX were added to the 10 µM NBQX-containing ACSF, kainate reduced the SR-induced Ca2+ signal by 25&20% (n=16). This effect of kainate/NBQX was statistically significantly smaller than that of kainate application in the absence of NBQX (P<0.05). Since a high density of presynaptic high-affinity kainate binding sites has been described on mossy fibre terminals,48 we also tested for effects of kainate in area CA3. Here, 90 min glucose deprivation resulted in irreversible loss of Schaffer collateralinduced responses, as well as in loss of signals induced by stimulation of the fimbria. Repetitive stimulation of the mossy fibres evoked decreases of (Ca2+)o. Bath applied kainate (5 µM) caused a reversible reduction of the mossy fibre stimulus-induced Ca2+ signal by 45&25% (n=6; Fig. 4), while 100 µM kainate reduced the Ca2+ signal by 50&33% (n=8). AMPA also suppressed the SR-induced (Ca2+)o decreases in area CA1 without effects on afferent volleys. In glucose deprivation-lesioned slices, 100 µM bath applied AMPA reduced the Ca2+ signal by 38&7% (n=6). Droplet application of 1 mM AMPA reduced the Ca2+ signal by 42&14% (n=22). Droplet application of 1 mM AMPA in the presence of 1 mM NBQX in the droplet and 10 µM NBQX in the bath reduced the SR-induced Ca2+ signal by 33&17% (n=33). This effect of AMPA was not statistically significantly different from that of AMPA applications in the absence of NBQX. NMDA decreased SR stimulus-induced Ca2+ signals in glucose deprivation-lesioned slices. Bath applied NMDA (50, 100 and 500 µM) reduced the Ca2+ signals reversibly by 54&14% (n=7), 54&15 (n=6) and 59&8% (n=4), respectively. NMDA (1 mM) applied in a droplet reduced the Ca2+ signals by 36&12% (n=25). When 2-APV was applied via the bath at a concentration of 30 µM, the amplitudes of the decreases in (Ca2+)o were comparable to those elicited in normal ACSF. When 1 mM NMDA was applied in a droplet also containing 1 mM 2-APV to the slice perfused with a 30 µM 2-APV-containing ACSF, the SR-induced Ca2+ signals were reduced by

103

Fig. 4. Effects of bath-applied kainate on presumed presynaptic Ca2+ entry in area CA3 in a slice lesioned by 90 min glucose deprivation. (Aa–c) After irreversible loss of postsynaptic signal components caused by glucose deprivation, 5 µM kainate depressed the mossy fibre stimulation (MFSTIM)-induced changes in (Ca2+)o significantly and reversibly. During application of kainate there was no change in baseline (Ca2+)o. Transient field potential responses are not shown, since in area CA3 mossy fibre stimulation did not evoke regular afferent volleys after irreversible loss of postsynaptic signal components. Horizontal bars indicate a stimulus train.

27&13% (n=40). This effect was significantly smaller than the effect of NMDA in the absence of 2-APV (P<0.05). Bath applied trans-ACPD (50 µM) had no significant effect (n=5) on SR stimulus-induced decreases in (Ca2+)o (Fig. 5). However, in four of five experiments, the overshoot in (Ca2+)o following the decrease was augmented by about 20%. In one experiment the overshoot was also prolonged. This effect did not reverse within 120 min after application of trans-ACPD. When glutamate was bath applied at concentrations of 10 mM in glucose deprivation-lesioned

104

K. Alici et al.

Fig. 5. Effects of bath-applied trans-ACPD on presumed presynaptic Ca2+ entry in area CA1 in a slice lesioned by 60 min glucose deprivation. (Aa, b) After irreversible loss of postsynaptic signal components caused by glucose deprivation, trans-ACPD applied at 50 µM had no effect on SR stimulation (SR-STIM)-induced changes in (Ca2+)o. During application of trans-ACPD there was no change in baseline (Ca2+)o or blockade of afferent volleys. Note that trans-ACPD augmented the Ca2+ overshoot following stimulus-induced (Ca2+)o decrease in this experiment. Asterisks indicate single stimuli and horizontal bars a stimulus train. Voltage calibration in Ab2 applies to Aa2.

slices, the SR-induced (Ca2+)o decreases were reduced by 60&17% (n=6). This effect was fully reversible. Droplet application of 100 mM glutamate caused a reversible reduction of the SR-induced decrease in (Ca2+)o by 80&22% (n=33). Bath application of 30 µM CNQX/30 µM 2-APV (n=14; Fig. 6), 10 µM NBQX (n=8) or 30 µM 2-APV (n=6) in kainate-lesioned and glucose deprivationlesioned slices had no significant effects on SRinduced decreases in (Ca2+)o. Also, droplet application of 1 mM 2-APV (n=11) or 1 mM NBQX (n=12) had no significant effects on SR-induced Ca2+ signals in lesioned slices.

Effects of glutamate receptor agonists on decreases in (Ca2+)o induced by repetitive stratum radiatum stimulation under conditions of blocked excitatory synaptic transmission The glutamatergic synaptic transmission in area CA1 was blocked by 30 µM CNQX and 30 µM 2-APV. Under this condition, SR stimulation elicited afferent volleys and by repetitive stimulation the largest (Ca2+)o decreases in the SR. These decreases are comparable to those after lesioning by glucose

deprivation and high kainate, and presumably also represent presynaptic Ca2+ influx (Fig. 1). Under conditions of blocked synaptic transmission, 100 µM bath-applied kainate reversibly reduced the SR stimulation-induced decreases of (Ca2+)o by 42&5% (n=3) without affecting the baseline (Ca2+)o and afferent volleys. Bath applied NMDA (100 µM) caused a reversible reduction of the SR stimulationinduced decreases of (Ca2+)o by 45&8% (n=4). Also, the NMDA application had no effects on baseline (Ca2+)o and afferent volleys. In contrast, kainate and NMDA had, in control experiments in the absence of CNQX and 2-APV, clear effects on the baseline (Ca2+)o and the baseline field potential. DISCUSSION

The present findings confirm that afferent fibre stimulation in the hippocampus leads to decreases in (Ca2+)o which persist to some degree when synaptic transmission is blocked.1,28,29,52,67 Similar decreases in (Ca2+)o are seen when the slices are treated with lesioning procedures such as high concentrations of kainate or by glucose deprivation. In these conditions we observed no changes in (Ca2+)o upon alvear

Effects of glutamate receptor agonists on Ca2+ uptake

105

Fig. 6. Effects of bath-applied CNQX/2-APV on presumed presynaptic Ca2+ entry in area CA1 in a slice lesioned by 90 min glucose deprivation. (Aa, b) After irreversible loss of postsynaptic signal components caused by glucose deprivation, 30 µM CNQX/30 µM 2-APV had no effect on SR stimulation (SR-STIM)induced changes in (Ca2+)o. During application of CNQX/2-APV there was no change in baseline (Ca2+)o or afferent volleys. Asterisks indicate single stimuli and horizontal bars a stimulus train. Voltage calibration in Ab2 applies to Aa2.

stimulation. Most fibres in the alveus originate from pyramidal cells. The disappearance of these signals suggests a rather complete damage of pyramidal cells. Also, the abolished transient field potential responses to alvear stimuli and the disappearance of postsynaptic components in the SR stimulus-induced transient field potential responses suggested a rather complete destruction of pyramidal cells. The persistence of afferent volleys and (Ca2+)o decreases in response to SR stimulation indicates that axons and presynaptic terminals remain functional for some time after lesioning of cell somata and dendrites. Stratum radiatum stimulation-induced decreases in (Ca2+)o in lesioned slices are presumably caused by Ca2+ entry into presynaptic terminals The decreases of (Ca2+)o observed after afferent stimulation in lesioned slices and in slices pretreated with CNQX and 2-APV presumably result from presynaptic Ca2+ entry.1,23,67 Comparison of extracellularly measured presumed presynaptic Ca2+ uptake with data where presynaptic Ca2+ uptake in area CA1 was monitored with Fura 2 show an extraordinary degree of congruence between the two methods. This applies to adenosine52,62,65 and

The K+ channel blocker baclofen.21,29,66 4-aminopyridine causes an increase of presynaptic Ca2+ influx in extracellular measurements of presynaptic signals25,51,52 and in Fura 2 measurements of presynaptic terminal Ca2+ signals.63 Nifedipine had no effect on presynaptic Ca2+ signals measured with Fura 263 and on stimulus-induced (Ca2+)o decreases studied under conditions of blocked synaptic transmission.24,30 Moreover, Ca2+ channel blockers such as Ni2+ or Cd2+ blocked extracellularly measured presynaptic Ca2+ signals24,25 and nearly completely depressed single terminal Ca2+ signals in area CA1.64 Also, effects of ù-agatoxin-IVA on presynaptic Ca2+ load63 and on stimulus-induced decreases in (Ca2+)o under conditions of blocked synaptic transmission23 are comparable, as well as the ineffectiveness of CNQX and 2-APV on these parameters.63–66 In contrast, postsynaptic increases of intracellular Ca2+ concentration were blocked by 2-APV and CNQX.37,47 From this we conclude that the stimulusinduced decreases of (Ca2+)o measured under conditions of blocked synaptic transmission or after lesioning of slices represent presynaptic Ca2+ uptake. It may be argued that glial cells may contribute to stimulus-induced changes in (Ca2+)o. Glial cells

106

K. Alici et al.

express glutamate receptors and these in turn can be Ca2+-permeable.5,7,8,35,36,46,58 Such a case has been reported for cerebellar7,8,36 and hippocampal glia46,58 in young animals which express Ca2+-permeable kainate, AMPA and NMDA receptors, which were antagonized by AMPA/kainate and NMDA receptor antagonists, respectively.7,8,36,46 The possibility therefore exists that glial cells activated by synaptically released glutamate contribute to the decreases in (Ca2+)o. While the reported glial signals were sensitive to the respective antagonists, neither CNQX, NBQX nor 2-APV blocked SR stimulus-induced decreases in (Ca2+)o in lesioned slices. Moreover, in contrast to intact slices, we did not note any detectable shifts in baseline (Ca2+)o during bath application of glutamate, kainate, NMDA, AMPA or transACPD in the lesioned slices of adult animals. Also, droplet application of these agonists in rather high concentrations left the baseline (Ca2+)o unaltered. We therefore assume that in adult hippocampal slices the density of glutamate receptors on glial cells is too low to produce any detectable Ca2+ entry from the extracellular space. The situation may be different in slices from young rats, where at least kainate could produce significant decreases in baseline (Ca2+)o in lesioned slices after complete loss of postsynaptic signal components (Alici K. and Heinemann U., unpublished observations). The stimulus-induced changes in (Ca2+)o observed in lesioned slices thus likely represent presynaptic Ca2+ entry.

Effects of glutamate receptor agonists and antagonists on presumed presynaptic Ca2+ signals Interestingly, all ionotropic glutamate receptor agonists and glutamate itself reversibly depressed SR stimulation-induced decreases in (Ca2+)o in area CA1 of lesioned slices, albeit only in relatively high concentrations. The highest concentration needed was that of glutamate. However, the equilibration timecourse of glutamate in a slice is already slow,18 and will be altered by high-affinity glutamate uptake.6,11,26,42 With droplet application we estimate that kainate, NMDA, AMPA, NBQX and 2-APV reach, at the site of recording, less than 10% of the applied concentration, which is less than 100 µM. Also, glutamate presumably does not reach much higher concentrations. Since bath-applied substances require more than 50 min for full equilibration in interface slice chambers,38 we believe that the relevant concentrations are probably less than 40–60% of the nominal bath concentration. This would suggest that rather high concentrations of kainate, NMDA, AMPA and presumably also glutamate are required to suppress presynaptic Ca2+ entry. Interestingly, CNQX, NBQX and 2-APV bath or droplet applied did not affect the stimulus-induced decreases in (Ca2+)o in lesioned slices, suggesting that either glutamate release no longer occurs or that the con-

centrations of glutamate reached during repetitive stimulation are too weak to activate presynaptic autoreceptors. The question arises as to which type of glutamate receptors are present on presynaptic endings. The fact that kainate and NMDA responses could be modified by NBQX and 2-APV, respectively, suggests that at least some of the binding sites at presynaptic endings respond to classical agonists and antagonists. A novel type of glutamate receptor involved in the control of presynaptic endings is, on the other hand, suggested by the fact that even high concentrations of NBQX, which has in binding studies a much higher affinity to AMPA than to kainate receptors,54 are without any statistically significant effect on the depressant effect of AMPA on presynaptic Ca2+ signals. Moreover, under conditions of blocked synaptic transmission by CNQX and 2-APV, application of kainate and NMDA still had a noticeable and rather large effect. Interestingly, a presynaptic glutamate receptor with a completely different structure compared with the known postsynaptic types has been described.56 When this receptor was expressed in Xenopus oocytes, it was pharmacologically but not electrophysiologically similar to NMDA receptors. Nevertheless, it is unknown which function this receptor possesses in the intact neuronal network. It could be that this receptor, when expressed in presynaptic endings, serves a regulatory control of presynaptic Ca2+ entry, thereby reducing too much glutamate release from presynaptic endings. With respect to metabotropic glutamate receptors, we have so far been unable to demonstrate any effects of trans-APCD except for an augmented Ca2+ overshoot following stimulus-induced (Ca2+)o reductions. These overshoots are presumably due to Ca2+ extrusion from presynaptic endings and may point to an augmented intracellular Ca2+ release.2 Baskys and Malenka4 showed that trans-ACPD depressed excitatory synaptic transmission, presumably via a presynaptic action. However, this effect was maximal during the first postnatal weeks and rather weak in adult slices. In agreement with our results, they showed that the presynaptic fibre volley was unaffacted by application of trans-ACPD. It has long been known that presynaptic fibres can express high-affinity kainate binding sites in area CA3.17,48 However, the consequences for presynaptic Ca2+ uptake are unclear. In area CA3, 5 µM kainate depressed the presynaptic Ca2+ signals, whereas in area CA1 kainate reduced the presynaptic Ca2+ influx only at concentrations higher than 50 µM, which confirmed the higher affinity of presynaptic kainate receptors on mossy fibre terminals. It would be interesting to learn more on the pharmacological properties of these kainate effects in area CA3. The mechanisms by which glutamate receptor agonists depress presynaptic Ca2+ entry are unclear.

Effects of glutamate receptor agonists on Ca2+ uptake

Most evidence suggests that predominantly P-type Ca2+ channels control presynaptic Ca2+ influx in glutamatergic nerve terminals.13,59,60 This is suggested by the fact that presynaptic Ca2+ signals measured under comparable conditions are sensitive to ù-agatoxin-IVA,23,63 which also very effectively suppresses postsynaptic signal components.12,13,31 The P-type Ca2+ currents show little time-dependent inactivation,60 and hence a depolarization by glutamate or glutamate receptor agonists should not lead to time-dependent inactivation of Ca2+ uptake. Unlike L-type Ca2+ channels, they are also not controlled by elevations of intracellular Ca2+.12,19 We therefore assume that the effects of glutamate and glutamate receptor agonists on presynaptic Ca2+ signals are not due to presynaptic intracellular Ca2+ accumulation. Although we did not find any significant effects on presynaptic volleys, we cannot exclude a depressant effect on presynaptic Ca2+ uptake by a prolonged presynaptic depolarization, similar to the mechanism discussed in primary afferent depolarizations in spinal cord preparations.14–16 There, it is assumed that presynaptic depolarization with a voltage-dependent inactivation of sodium channels results in a reduced electrotonic propagation of action potentials into the presynaptic terminals, thereby opening less Ca2+ channels. Of course, effects on

107

presynaptic Ca2+ channels, K+ channels or a second messenger system cannot be excluded.

CONCLUSION

The fact that NMDA, kainate, AMPA and glutamate reduced presynaptic Ca2+ entry might be a mechanism to prevent excessive glutamate release. Indeed, perfusion measurements during ischaemia showed a transient increase in glutamate which reversed temporarily during continued ischaemia.49,61 This might be due in part to glutamate uptake, but also to blockade of further glutamate release by presynaptic autoreceptors. The fact that inhibitory presynaptic autoreceptors exist for glutamate is of considerable interest. Such receptors might be the ideal target site for new types of anticonvulsants. Drugs which augment the effects of intrinsically released glutamate on these receptors might be of special interest in this regard. Acknowledgements—This research was supported by a grant from the DFG (He1128/6-2), the SFB 194, the HFSP and the BMFT. We are indebted to M. Bullmann and A. Du¨erkop for excellent technical assistance. We thank Drs W. Mu¨ller and M. Numberger for critical reading of earlier versions of the manuscript.

REFERENCES

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.

Alici K. and Heinemann U. (1995) Effects of low glucose levels on changes in (Ca2+)o induced by stimulation of Schaffer collaterals under conditions of blocked chemical synaptic transmission in rat hippocampal slices. Neurosci. Lett. 185, 5–8. Arens J., Stabel J. and Heinemann U. (1992) Pharmacological properties of excitatory amino acid induced changes in extracellular calcium concentration in rat hippocampal slices. Can. J. Physiol. Pharmac. 70 Suppl., 194–205. Auer R., Kalimo H., Olsson Y. and Wieloch T. (1985) The dentate gyrus in hypoglycemia: pathology implicating excitotoxin-mediated neuronal necrosis. Acta neuropath., Berlin 67, 279–288. Baskys A. and Malenka R. C. (1991) Agonists at metabotropic glutamate receptors presynaptically inhibit EPSCs in neonatal rat hippocampus. J. Physiol. 444, 687–701. Berger T. (1995) AMPA-type glutamate receptors in glial precursor cells of the rat corpus callosum: ionic and pharmacological properties. Glia 14, 101–114. Bouvier M., Szatkowski M., Amato A. and Attwell D. (1992) The glial cell glutamate uptake carrier countertransports pH-changing anions. Nature 360, 471–474. Brune T. and Deitmer J. W. (1995) Intracellular acidification and Ca2+ transients in cultured rat cerebellar astrocytes evoked by glutamate agonists and noradrenaline. Glia 14, 153–161. Burnashev N., Khodorova A., Jonas P., Helm P. J., Wisden W., Monyer H., Seeburg P. H. and Sakmann B. (1992) Calcium-permeable AMPA–kainate receptors in fusiform cerebellar glial cells. Science 256, 1566–1570. Chesselet M.-F. (1984) Presynaptic regulation of neurotransmitter release in the brain: facts and hypothesis. Neuroscience 12, 347–375. Choi D. W. (1992) Excitotoxic cell death. J. Neurobiol. 23, 1261–1276. Danbolt N. C. (1994) The high affinity uptake system for excitatory amino acids in the brain. Prog. Neurobiol. 44, 377–396. De Leon M., Wang Y., Jones L., Perez-Reyes E., Wei X., Soong T. W., Snutch T. P. and Yue D. T. (1995) Essential Ca2+-binding motif for Ca2+-sensitive inactivation of L-type Ca2+ channels. Science 270, 1502–1506. Dunlap K., Luebke J. I. and Turner T. J. (1995) Exocytotic Ca2+ channels in mammalian central neurons. Trends Neurosci. 18, 89–98. Eccles J. C., Eccles R. M. and Magni F. (1961) Central inhibitory action attributable to presynaptic depolarization produced by muscle afferent volleys. J. Physiol. 159, 147–166. Eccles J. C., Magni F. and Willis W. D. (1962) Depolarization of central terminals of group I afferent fibres from muscle. J. Physiol. 160, 62–93. Eccles J. C., Schmidt R. F. and Willis W. D. (1962) Presynaptic inhibition of the spinal monosynaptic reflex pathway. J. Physiol. 161, 282–297. Foster A. C., Mena E. E., Monaghan D. T. and Cotman C. W. (1981) Synaptic localization of kainic acid binding sites. Nature 289, 73–75. Garthwaite G., Williams G. D. and Garthwaite J. (1992) Glutamate toxicity: an experimental and theoretical analysis. Eur. J. Neurosci. 4, 353–360.

108 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49.

50. 51. 52. 53. 54.

K. Alici et al. Haack J. A. and Rosenberg R. L. (1994) Calcium-dependent inactivation of L-type calcium channels in planar lipid bilayers. Biophys. J. 66, 1051–1060. Heinemann U., Albrecht D., Ko¨hr G., Rausche G., Stabel J. and Wisskirchen T. (1992) Low-Ca2+-induced epileptiform activity in rat hippocampal slices. Epilepsy Res. Suppl. 8, 147–155. Heinemann U., Hamon B. and Konnerth A. (1984) GABA and baclofen reduce changes in extracellular free calcium in area CA1 of rat hippocampal slices. Neurosci. Lett. 47, 295–300. Heinemann U., Lux H. D. and Gutnick M. J. (1977) Extracellular free calcium and potassium during paroxysmal activity in the cerebral cortex of the cat. Expl Brain Res. 27, 237–243. Igelmund P., Zhao Y. Q. and Heinemann U. (1996) Effects of T-, L-, N-, and P-type calcium channel blockers on stimulus-induced pre- and postsynaptic calcium fluxes in rat hippocampal slices. Expl Brain Res. Jones R. S. G. and Heinemann U. (1987) Differential effects of calcium entry blockers on pre- and postsynaptic influx of calcium in the rat hippocampus in vitro. Brain Res. 416, 257–266. Jones R. S. G. and Heinemann U. (1987) Pre- and postsynaptic K+ and Ca2+ fluxes in area CA1 of the rat hippocampus in vitro: effects of Ni2+, TEA and 4-AP. Expl Brain Res. 68, 205–209. Kanai Y., Smith C. P. and Hediger M. A. (1993) The elusive transporters with a high affinity for glutamate. Trends Neurosci. 16, 365–370. Ko¨hler C. and Schwarcz R. (1983) Comparison of ibotenate and kainate neurotoxicity in rat brain: a histological study. Neuroscience 8, 819–835. Konnerth A. and Heinemann U. (1983) Effects of GABA on presumed presynaptic Ca2+ entry in hippocampal slices. Brain Res. 270, 185–189. Leweke F. M., Rausche G. and Heinemann U. (1990) Proconvulsive effects of baclofen in dentate gyrus are not related to effects of baclofen on presynaptic calcium entry. Neurosci. Res. Commun. 6, 83–88. Louvel J., Abbes S. and Godfraind J. M. (1986) Effect of organic calcium channel blockers on neuronal calcium-dependent processes. In Calcium Electrogenesis and Neuronal Functioning (eds Heinemann U., Klee M., Neher E. and Singer W.), pp. 375–385. Springer, Berlin. Luebke J. I., Dunlap K. and Turner T. J. (1993) Multiple calcium channel types control glutamatergic synaptic transmission in the hippocampus. Neuron 11, 895–902. Lux H. D. and Neher E. (1973) The equilibration time course of (K+)o in cat cortex. Expl Brain Res. 17, 190–205. Miwa A., Robinson H. P. C. and Kawai N. (1993) Presynaptic glutamate receptors depress inhibitory postsynaptic transmission in lobster neuromuscular synapse. J. Neurophysiol. 70, 1159–1167. Mody I., Lambert J. D. C. and Heinemann U. (1987) Low extracellular magnesium induces epileptiform activity and spreading depression in rat hippocampal slices. J. Neurophysiol. 57, 869–888. Mu¨ller T., Grosche J., Ohlemeyer C. and Kettenmann H. (1993) NMDA-activated currents in Bergmann glial cells. NeuroReport 4, 671–674. Mu¨ller T., Mo¨ller T., Berger T., Schnitzer J. and Kettenmann H. (1992) Calcium entry through kainate receptors and resulting potassium-channel blockade in Bergmann glial cells. Science 256, 1563–1566. Mu¨ller W. and Connor J. A. (1991) Dendritic spines as individual neuronal compartments for synaptic Ca2+ responses. Nature 354, 73–76. Mu¨ller W., Misgeld U. and Heinemann U. (1988) Carbachol effects on hippocampal neurons in vitro: dependence on the rate of rise of carbachol tissue concentration. Expl Brain Res. 72, 287–298. Nadler J. V., Evenson D. A. and Cuthbertson G. J. (1981) Comparative toxicity of kainic acid and other acidic amino acids toward rat hippocampal neurons. Neuroscience 6, 2505–2517. Nadler J. V., Perry B. W. and Cotman C. W. (1978) Intraventricular kainic acid preferentially destroys hippocampal pyramidal cells. Nature 271, 676–677. Nakanishi S. (1992) Molecular diversity of glutamate receptors and implication for brain function. Science 258, 597–603. Nicholls D. and Attwell D. (1990) The release and uptake of excitatory amino acids. Trends pharmac. Sci. 11, 462–468. Nicholls D. G. (1993) The glutamatergic nerve terminal. Eur. J. Biochem. 212, 613–631. Oka A., Belliveau M. J., Rosenberg P. A. and Volpe J. J. (1993) Vulnerability of oligodendroglia to glutamate: pharmacology, mechanisms, and prevention. J. Neurosci. 13, 1441–1453. Olney J. W., Fuller T. and De Gubareff T. (1979) Acute dendrotoxic changes in the hippocampus of kainate treated rats. Brain Res. 176, 91–100. Porter J. T. and McCarthy K. D. (1995) GFAP-positive hippocampal astrocytes in situ respond to glutamatergic neuroligands with increases in (Ca2+)i. Glia 13, 101–112. Regehr W. G. and Tank D. W. (1991) Selective fura-2 loading of presynaptic terminals and nerve cell processes by local perfusion in mammalian brain slice. J. Neurosci. Meth. 37, 111–119. Represa A., Tremblay E. and Ben-Ari Y. (1987) Kainate binding sites in the hippocampal mossy fibers: localization and plasticity. Neuroscience 20, 739–748. Scheller D., Kolb J., Szathmary S., Zacharias E., de Ryck M., v. Reempts J., Clinicke G. and Tegtmeier F. (1995) XLUI-15. Extracellular changes of glutamate in the periinfarct zone. Effect of Lubeluzole. J. cerebr. Blood Flow Metab. 15 Suppl. 1, 379. Schoepp D. D. (1994) Novel functions for subtypes of metabotropic glutamate receptors. Neurochem. Int. 24, 439–449. Schubert P. and Heinemann U. (1988) Adenosine antagonists combined with 4-aminopyridine cause partial recovery of synaptic transmission in low Ca media. Expl Brain Res. 70, 539–549. Schubert P., Heinemann U. and Kolb R. (1986) Differential effects of adenosine on pre- and postsynaptic calcium fluxes. Brain Res. 376, 382–386. Schwarcz R., Whetsell W. O. and Mangano R. M. (1983) Quinolinic acid: an endogenous metabolite that produces axon sparing lesions in rat brain. Science 219, 316–318. Sheardown M. J., Nielsen E. O., Hansen A. J., Jacobsen P. and Honore´ T. (1990) 2,3-Dihydroxy-6-nitro-7-sulfamoylbenzo(F)quinoxaline: a neuroprotectant for cerebral ischemia. Science 247, 571–574.

Effects of glutamate receptor agonists on Ca2+ uptake 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67.

109

Siesjo¨ B. K. and Bengtsson F. (1989) Calcium fluxes, calcium antagonists, and calcium-related pathology in brain ischemia, hypoglycemia, and spreading depression: a unifying hypothesis. J. cerebr. Blood Flow Metab. 9, 127–140. Smirnova T., Stinnakre J. and Mallet J. (1993) Characterization of a presynaptic glutamate receptor. Science 262, 430–433. Stanton P. K., Mody I. and Heinemann U. (1989) A role for N-methyl--aspartate receptors in norepinephrineinduced long-lasting potentiation in the dentate gyrus. Expl Brain Res. 77, 517–530. Steinha¨user C., Jabs R. and Kettenmann H. (1994) Properties of GABA and glutamate responses in identified glial cells of the mouse hippocampal slice. Hippocampus 4, 19–36. Turner T. J., Adams M. E. and Dunlap K. (1992) Calcium channels coupled to glutamate release identified by ù-Aga-IVA. Science 258, 310–313. Uchitel O. D. and Protti D. A. (1995) P-type calcium channels and transmitter release from nerve terminals. NIPS 9, 101–105. Wahl F., Obrenovitch T. P., Hardy A. M., Plotkine M., Boulu R. and Symon L. (1994) Extracellular glutamate during focal cerebral ischaemia in rats: time course and calcium dependency. J. Neurochem. 63, 1003–1011. Wu L.-G. and Saggau P. (1993) Adenosine inhibits synaptic transmission by reducing the presynaptic calcium influx at CA3–CA1 synapses of guinea pig hippocampus. Soc. Neurosci. Abstr. 19, 1518. Wu L.-G. and Saggau P. (1994) Pharmacological identification of two types of presynaptic voltage-dependent calcium channels at CA3–CA1 synapses of the hippocampus. J. Neurosci. 14, 5613–5622. Wu L.-G. and Saggau P. (1994) Presynaptic calcium is increased during normal synaptic transmission and paired-pulse facilitation, but not in long-term potentiation in area CA1 of hippocampus. J. Neurosci. 14, 645–654. Wu L.-G. and Saggau P. (1994) Adenosine inhibits evoked synaptic transmission primarily by reducing presynaptic calcium influx in area CA1 of hippocampus. Neuron 12, 1139–1148. Wu L.-G. and Saggau P. (1995) GABAB receptor-mediated presynaptic inhibition in guinea-pig hippocampus is caused by reduction of presynaptic Ca2+ influx. J. Physiol. 485, 649–657. Young J. N. and Somjen G. G. (1992) Suppression of presynaptic calcium currents by hypoxia in hippocampal tissue slices. Brain Res. 573, 70–76. (Accepted 15 July 1996)