- Email: [email protected]
Tonic facilitation of glutamate release by presynaptic N-methyl-d -aspartate autoreceptors in the entorhinal cortex
Pergamon
PII:
Neuroscience Vol. 75, No. 2, pp. 339–344, 1996 Copyright ? 1996 IBRO. Published by Elsevier Science Ltd Printed in Great Britain 0306–4522/96 $15.00+0.00 S0306-4522(96)00301-6
Letter to Neuroscience TONIC FACILITATION OF GLUTAMATE RELEASE BY PRESYNAPTIC N-METHYL--ASPARTATE AUTORECEPTORS IN THE ENTORHINAL CORTEX N. BERRETTA* and R. S. G. JONES† Department of Pharmacology, University of Oxford, Mansfield Road, Oxford OX1 3QT, U.K. Key words: entorhinal cortex, NMDA, presynaptic receptors, glutamate release.
N-Methyl-D-aspartate receptors are fundamental for neuronal plasticity and development in the CNS.3,22 Most studies have examined postsynaptic roles of this receptor, but there are also indications for a presynaptic location and function.6,18,23 Here, we provide electrophysiological evidence for the existence of presynaptic N-methyl-D-aspartate receptors which can tonically facilitate glutamate release in the CNS. The N-methyl-D-aspartate receptor antagonist 2-amino-5phosphonopentanoate reduced the frequency, but not amplitude, of glutamate-mediated spontaneous excitatory postsynaptic currents in layer II neurons of the rat entorhinal cortex. This effect was also observed in the presence of tetrodotoxin and when postsynaptic N-methyl-D-aspartate receptors were blocked by dialysis with dizocilpine maleate. When extracellular calcium was replaced with strontium, 2-amino5-phosphonopentanoate reduced the ‘‘tail’’ of spontaneous excitatory postsynaptic currents that followed an evoked excitatory postsynaptic current. Finally, there was a tendency for paired-pulse facilitation of excitatory postsynaptic currents evoked at short (50 ms) intervals with postsynaptic N-methyl-Daspartate receptors blocked) to be reduced by 2-amino5-phosphonopentanoate, although this did not reach significance. These data strongly support the presence of presynaptic N-methyl-D-aspartate autoreceptors *Present address: SISSA—ISAS International School for Advanced Studies, Via Beirut 2/4, 34014, Trieste, Italy. †To whom correspondence should be addressed. Abbreviations: AQE, asynchronous quantal event; BAPTA, 1,2-bis(2-aminophenoxy)ethane-N,N,N*,N*-tetraacetic acid; -AP5, 2-amino-5-phosphonopentanoate; EC, entorhinal cortex; EGTA, ethyleneglycolbis(aminoethyl ether)tetra-acetate; EPSC, excitatory postsynaptic current; HEPES, N-2-hydroxyethylpiperazine-N*-2-ethanesulphonic acid; mEPSC, miniature EPSC; MK-801, dizocilpine maleate; NMDA, N-methyl--aspartate; PPF, paired-pulse facilitation; QX-314, N-(2,6-dimethyl phenylcarbamoylmethyl) triethylammonium bromide; sEPSC, spontaneous EPSC.
which may facilitate glutamate release in layer II of the entorhinal cortex. Copyright ? 1996 IBRO. Published by Elsevier Science Ltd. We have recently shown that neurons of the rat entorhinal cortex (EC) display continuous, spontaneous synaptic excitation mediated by release of glutamate, acting primarily at non-N-methyl-aspartate (NMDA) receptors.2 Application of the NMDA receptor antagonist 2-amino-5-phosphonopentanoate (-AP5; 100 µM, n = 6) did not affect the mean amplitude of spontaneous excitatory postsynaptic currents (sEPSCs) in these neurons (9.5 & 0.1 and 9.1 & 0.1 pA, before and after -AP5). Because the amplitudes of these events are not normally distributed,2 a non-parametric comparison of cumulative probability distributions was performed using the Kolmogorov–Smirnov test.24 This confirmed the lack of effect on amplitude of sEPSCs. However, a similar comparison of frequency of these events showed a clear reduction (P < 0.001, Kolmogorov–Smirnov test) in the presence of the antagonist (see Fig. 1). The mean interval between events increased significantly, from 135.1 & 2.7 to 194.8 & 5.4 ms, in the presence of -AP5. We have considered a number of interpretations of our observations. (1) -AP5 could cause a reduction of activity-dependent sEPSCs, normally driven by NMDA receptors in excitatory networks within the EC. (2) Postsynaptic NMDA receptors could induce an increase in presynaptic release via production of a retrograde messenger.4 (3) The frequency of sEPSCs could be set by a tonic feedback facilitatory action of glutamate on presynaptic NMDA receptors. To test the first two possibilities, we studied the effect of -AP5 on activity-independent, miniature excitatory postsynaptic currents (mEPSCs) recorded in the presence of tetrodotoxin (1 µM). To obviate a role of postsynaptic NMDA receptors, we included the open-channel NMDA receptor blocker,
339
340
N. Berretta and R. S. G. Jones
Fig. 1. -AP5 reduces the frequency but not amplitude of sEPSCs in layer II. (a) The traces in the upper panels are consecutive records of spontaneous activity recorded in a layer II neuron in the absence (left) and presence (right) of -AP5 (100 µM). The NMDA antagonist reduced the frequency of sEPSCs. The graphs show the cumulative distributions of amplitude and interval obtained by collecting 280 sEPSCs before (solid line) and after (dashed line) -AP5, from the same neuron. They show a significant (P < 0.001) increase in interval between events, with no change in amplitude. (b) Pooled data obtained by collecting 100 events from each of six neurons before (solid line) and after (dashed line) -AP5. Again, the interval but not the amplitude distribution was significantly (P < 0.001) altered by -AP5. Experiments were performed in slices (450 µm) of the EC, prepared from Wistar rats (120–150 g) as described previously.16 They were maintained at 34 & 0.2)C between a continuous (1.2 ml/min) perfusion with artificial cerebrospinal fluid and warmed, humidified carbogen gas. The artificial cerebrospinal fluid contained (in mM): NaCl 126, KCl 3, NaHCO3 24, NaH2PO4 1.25, MgSO4 2, CaCl2 2, -glucose 10. -AP5 (100 µM) was applied by bath perfusion or as a 5-µl droplet directly on to the slice. Whole-cell recordings were obtained using the ‘‘blind-patch’’ technique with glass electrodes filled with (in mM): CsMeSO4 130, HEPES 5, EGTA 0.5, MgCl2 1, NaCl 1, CaCl2 0.34, QX-314 5 (pH 7.3–7.4). Currents were recorded at a holding potential of "70 mV using an Axopatch 200A amplifier. They were filtered at 2 kHz, digitized at 48 kHz and recorded on a DAT recorder. Compensation for series resistance was not employed and statistical comparisons were only made when series resistance did not change by more than 10%. Spontaneous events were detected automatically using a threshold crossing criterion. The threshold level varied from neuron to neuron but was always the same before and after drug application. Peak amplitude and interval distributions were compared using the K-S test. Synaptic responses were also evoked by electrical stimulation via a bipolar platinum stimulating electrode placed on to the surface of the slice in the deep layers (V and VI) of the EC. Data are expressed as mean & S.E.M.
Glutamate release by NMDA autoreceptors in the EC
341
Fig. 2. Blockade of postsynaptic NMDA receptors does not affect the reduction in frequency of sEPCs induced by -AP5. (a) The records in the upper line show electrically evoked synaptic responses (stimulus at arrowheads) recorded in a layer II neuron extensively dialysed without MK-801 in the filling solution. The fast excitatory postsynaptic current (EPSC) was followed by a fast GABAA receptor-dependent inhibitory postsynaptic current. Following perfusion with 6-nitro-7-sulphoamoylbenzo(f)-quinoxaline-2,3dione (10 µM) and bicuculline methochloride (10 µM) a slow EPSC remained. This increased dramatically in magnesium-free medium and was abolished by -AP5 (100 µM), showing that it was mediated by NMDA receptors. The lower line shows a different neuron dialysed with a filling solution containing MK-801 (1 mM). An inhibitory postsynaptic current was not detectable in this neuron and, following perfusion with 6-nitro-7-sulphoamoylbenzo(f)-quinoxaline-2,3-dione plus bicuculline methochloride, there was clearly no slow EPSC, even when magnesium was removed from the perfusion medium. This neuron was also recorded at a holding potential of "40 mV to maximize any contribution from NMDA receptors. Thus, it seems highly probable that the postsynaptic NMDA receptors of the recorded neuron were blocked by internal application of MK-801. (b) The cumulative distributions of peak amplitude and intervals of sEPSC (n = 400) events before (solid line) and after (dashed line) perfusion with -AP5 (100 µM), from a neuron recorded in the presence of tetrodotoxin (1 µM) and bicuculline methochloride (10 µM) and with MK-801 in the patch pipette. -AP5 still significantly (P < 0.001) increased the interval between sEPSCs, without a change in amplitude. In these experiments the filling solution contained (in mM): CsMeSO4 122, CsCl 3, HEPES 5, BAPTA 10, (+)-MK-801 1, QX-314 5 (pH 7.3–7.4).
dizocilpine maleate (MK-801; 1 mM), in the recording pipette. Control studies clearly showed that when cells were dialysed with MK-801, the NMDA receptor-mediated component of electrically evoked synaptic responses was completely absent (n = 10; Fig. 2a). However, when -AP5 was applied in the presence of tetrodotoxin plus intracellular MK-801, it still caused a significant increase in the interval between mEPSCs, from 138 & 5 to 169 & 8 ms (P < 0.05, n = 9). Again, the amplitude of the same events was unaffected (11.1 & 0.5 and 11.3 &
0.6 pA, before and after -AP5; Fig. 2b). Thus, we believe that a likely explanation for the reduction in frequency by -AP5 is blockade of presynaptic NMDA receptors which tonically facilitate glutamate release. Substitution of extracellular calcium with strontium has been shown to facilitate ‘‘asynchronous’’ release of neurotransmitter. Thus, in the presence of strontium, stimulation of afferent axons elicits a reduced synaptic response succeeded by a series of asynchronous quantal events (AQEs), which result
342
N. Berretta and R. S. G. Jones
Fig. 3. -AP5 reduces the number of AQEs following an evoked response in the presence of strontium. In these experiments, CaCl2 was replaced with SrCl2 (2 mM) in the perfusion medium. MK-801 was included in the pipette solution. The traces in the upper panel show EPSCs evoked by electrical stimulation at 0.1 Hz. The point of stimulation is marked by the arrowheads. The five consecutive responses superimposed on the top line show that the initial evoked response was followed by a series of AQEs. The lower line shows that -AP5 (100 µM) had no effect on the initial evoked response, but reduced the occurrence of the succeeding AQEs. Results are quantified in Table 1. Table 1. Pooled data from six neurons recorded with strontium substituted for calcium, showing the amplitude (upper line) and number (lower line) of asynchronous quantal events detected in 40 successive traces within 300 ms of the peak of the evoked response
Amplitude (pA) Number
Control
-AP5
10.9 & 0.4 4.7 & 0.7
10.5 & 0.3 3.3 & 0.6
The number of events was significantly reduced in -AP5 (P < 0.02, t-test paired data), but their mean amplitude was unaltered.
from transmitter release from the recently activated synapses.12,21 Such an effect was seen in layer II neurons and permitted us to examine transmitter release occurring after stimulation of the afferent pathway. Taking into consideration the fact that NMDA receptors are permeable to strontium,19 we have addressed the possibility that presynaptic NMDA receptors might modulate this effect. Neurons in this study were again recorded with MK-801 in the patch pipettes. The number of sEPSCs recorded in a period of 300 ms following the peak of the evoked response was decreased significantly by -AP5 (100 µM; Fig. 3). However, the mean amplitude of the events that remained was unchanged.
These data are summarized in Table 1 and suggest that NMDA receptors could presynaptically facilitate release in the previously activated afferents, perhaps as a result of an increase in calcium influx into the terminals. Paired-pulse facilitation (PPF) is also considered to be a presynaptic phenomenon, resulting from a transient increase of presynaptic calcium, caused by a conditioning synaptic response.13,27 We have examined whether presynaptic NMDA receptors might contribute to PPF. PPF was evident 20–200 ms after the conditioning response in neurons (n = 6) recorded with MK-801 in the patch pipettes (Fig. 4). Perfusion with -AP5 reduced the degree of facilitation, particularly at the 50 ms interval (Fig. 4). Although the change was not statistically significant, the reduction of the facilitation by -AP5 was consistent in every cell and restricted to the 50 ms interval. It is quite likely that PPF would occur as a result of calcium entry via both presynaptic NMDA receptors and voltage-gated calcium channels, and that the contribution from the latter is greater for activitydependent responses. Thus, the present results are consistent with the presence of presynaptic NMDA autoreceptors on glutamate terminals in layer II of the EC. We cannot definitely rule out the possibility that the receptors are subsynaptic at axoaxonic glutamate synapses on
Glutamate release by NMDA autoreceptors in the EC
Fig. 4. -AP5 can reduce PPF in neurons where postsynaptic NMDA receptors were blocked. (a) The panel on the left shows the superimposed evoked EPSCs in a layer II neuron with inter-pulse intervals of 50 and 200 ms. The point of stimulation is marked by the arrowheads. Each trace is the average of five sweeps. The right-hand traces show that -AP5 (100 µM) reduced the degree of PPF at 50 ms but not at 200 ms. (b) Pooled data from six neurons show a trend towards a reduction in facilitation, particularly at 50 ms, although overall this effect just failed to reach significance. All experiments included MK-801 in the patch pipette filling solutions.
the glutamate terminals, but to our knowledge such synapses have not been demonstrated in the EC. NMDA autoreceptors appear to be activated by spontaneously released glutamate and determine, by positive feedback, a tonic facilitation of glutamate release. Binding sites for NMDA on presynaptic terminals have been demonstrated immunocytochemically.23 A facilitatory effect of NMDA on monoamine release has been shown,10,17,20,26 and on the release of acetylcholine at developing neuromuscular synapses.11 Neurochemical experiments have also indicated a presynaptic facilitation of glutamate release by NMDA receptors.6,18 However, electrophysiological evidence for presynaptic autoreceptors has previously been lacking. Autoreceptor modulation of glutamate release has largely been considered a domain of metabotropic receptors,1,8,25 although presynaptic kainate receptors have also been reported to be involved in control of glutamate
343
release.7,9 Our results show that NMDA receptors may be positively involved in such modulation as well. Presently, we can only speculate on the mechanism of the potentiation of release by presynaptic NMDA receptors. Perhaps the most likely explanation is that presynaptic NMDA receptors increase calcium influx via the receptor-gated ionophore. We need to explore this possibility. Another question which we have to ask is whether transmitter release at all glutamate synapses is modulated by presynaptic NMDA receptors, and at present we think this is unlikely to be the case. In layer V neurons, there is a greater contribution of postsynaptic NMDA receptors to the amplitude and duration of sEPSCs than in layer II,2 but we have found little effect of -AP5 on their frequency (Berretta and Jones, unpublished). In CA1 neurons, Oliet et al.21 have reported that long-term potentiation is associated with an increase in amplitude and frequency of AQEs recorded in strontium and the reverse effect occurs with long-term depression. Both effects are blocked by -AP5, but the authors do not indicate whether the antagonist had any effect on the frequency of AQEs per se. The functional role of presynaptic NMDA receptors is also purely speculative at present. Repetitive activation of afferents to layer II results in a shortterm, frequency-dependent enhancement of glutamate transmission,15 and it has been suggested that this may be involved in processing of information destined for the hippocampus.14 Presynaptic NMDA receptors could play a role in this short-term enhancement of transmission in the EC. Also, there is a strong school of thought which suggests that an increase in glutamate release is responsible, in part, for the phenomenon of long-term potentiation.4 Although this is suggested to occur subsequent to postsynaptic release of a retrograde messenger, it would not be unreasonable to suggest that increased release could also involve long-term presynaptic changes at an NMDA autoreceptor. Finally, it is noteworthy that layer II of the EC suffers very early and severe neurodegeneration in Alzheimer’s disease,5 and it could be speculated that this might involve an over-stimulation of glutamate release as a result of increased activity of presynaptic NMDA receptors.
Acknowledgements—We thank the Wellcome Trust and Royal Society for financial support, P. Vincent for the software for the analysis program and C. Stoub for the Kolmogorov–Smirnov test.
REFERENCES
1. 2.
Baskys A. and Malenka R. C. (1991) Agonists of metabotropic glutamate receptors presynaptically inhibit EPSCs in neonatal rat hippocampus. J. Physiol., Lond. 444, 687–701. Berretta N. and Jones R. S. G. (1996) A comparison of spontaneous EPSCs in layer II and layer IV–V neurones of the rat entorhinal cortex in vitro. J. Neurophysiol. 76, 1–12.
344 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27.
N. Berretta and R. S. G. Jones Blanton M. G., LoTurco J. J. and Kriegstein A. R. (1989) Whole-cell recording from neurones in slices of reptilian and mammalian cerebral cortex. J. Neurosci. Meth. 30, 203–210. Bliss T. V. P. and Collingridge G. L. (1993) A synaptic model of memory: long-term potentiation in the hippocampus. Nature 361, 31–39. Braak H. and Braak E. (1991) Neuropathological staging of Alzheimer-related changes. Acta neuropath. 82, 239–259. Bustos G., Abarca J., Forray M. I., Gysfing K., Bradberry C. W. and Roth R. H. (1992) Regulation of excitatory amino acid release by N-methyl--aspartate receptors in rat striatum: in vivo microdialysis studies. Brain Res. 585, 105–115. Chittajallu R., Vignes M., Dev K. K., Barnes J. M., Collingridge G. L. and Henley J. M. (1996) Regulation of glutamate release by presynaptic kainate receptors in the hippocampus. Nature 379, 78–81. Desai C. H., Smith T. S. and Conn P. J. (1992) Multiple metabotropic glutamate receptors regulate hippocampal function. Synapse 12, 206–213. Ferkany J. W., Zaczek R. and Coyle J. T. (1982) Kainic acid stimulates excitatory amino acid neurotransmitter release at presynaptic receptors. Nature 298, 757–759. Fink K., Bo¨nisch H. and Gothert M. (1990) Presynaptic NMDA receptors stimulate noradrenaline release in the cerebral cortex. Eur. J. Pharmac. 185, 115–117. Fu W.-M., Liou J.-C., Lee Y.-H. and Liou H.-C. (1995) Potentiation of neurotransmitter release by activation of presynaptic glutamate receptors at developing neuromuscular synapses of Xenopus. J. Physiol., Lond. 489, 813–823. Goda Y. and Stevens C. F. (1994) Two components of transmitter release at a central synapse. Proc. natn. Acad. Sci. U.S.A. 91, 12942–12946. Hess G., Kuhnt U. and Voronin L. L. (1987) Quantal analysis of paired-pulse facilitation in guinea pig hippocampal slices. Neurosci. Lett. 77, 187–192. Jones R. S. G. (1993) Entorhinal–hippocampal connections: a speculative view of their functions. Trends Neurosci. 16, 58–64. Jones R. S. G. (1994) Synaptic and intrinsic properties of neurones of origin of the perforant path in layer II of the rat entorhinal cortex in vitro. Hippocampus 4, 335–353. Jones R. S. G. and Heinemann U. (1988) Synaptic and intrinsic responses of medial entorhinal cortical cells in normal and magnesium-free medium in vitro. J. Neurophysiol. 59, 1476–1496. Krebs M. O., Desce J. M., Kemel M. L., Gauchy C., Godeheu G., Cheramy A. and Glowinski J. (1991) Glutamatergic control of dopamine release in the rat striatum: evidence for presynaptic N-methyl--aspartate receptors on dopaminergic nerve terminals. J. Neurochem. 56, 81–85. Martin D., Bustos G. A., Bowe M. A., Bray S. D. and Nadler J. V. (1991) Autoreceptor regulation of glutamate and aspartate release from slices of the hippocampal CA1 area. J. Neurochem. 56, 1647–1655. Mayer M. L. and Westbrook G. L. (1987) Permeation and block of N-methyl--aspartic acid receptor channels by divalent cations in mouse cultured central neurones. J. Physiol., Lond. 394, 501–527. Moghaddam B., Gruen R. J., Roth R. H., Bunney B. S. and Adams R. N. (1990) Effect of -glutamate on the release of striatal dopamine: in vivo dialysis and electrochemical studies. Brain Res. 518, 55–60. Oliet S. H. R., Malenka R. C. and Nicoll R. A. (1995) Bidirectional control of quantal size by synaptic activity in the hippocampus. Science 271, 1294–1297. Scheetz A. J. and Constantine-Paton M. (1994) The modulation of NMDA receptor function: implications for vertebrate neural development. Fedn Proc. Fedn Am. Socs exp. Biol. 8, 745–752. Siegel S. J., Brose N., Janssen W. G., Gasic G. P., Jahn R., Heinemann S. F. and Morrison J. H. (1994) Regional, cellular, and ultrastructural distribution of N-methyl--aspartate receptor subunit 1 in monkey hippocampus. Proc. natn. Acad. Sci. U.S.A. 91, 564–568. Van de Kloot W. (1991) The regulation of quantal size. Prog. Neurobiol. 36, 93–130. Vignes M., Clarke V. R. J., Davies C. H., Chambers A., Jane D. E., Watkins J. C. and Collingridge G. L. (1995) Pharmacological evidence for an involvement of group II and group III mGluRs in the presynaptic regulation of excitatory synaptic responses in the CA1 region of rat hippocampal slices. Neuropharmacology 34, 973–982. Wang J. K. T., Andrews H. and Thukral V. (1992) Presynaptic glutamate receptors regulate noradrenaline release from isolated nerve terminals. J. Neurochem. 58, 204–211. Zucker R. S. (1989) Short-term synaptic plasticity. A. Rev. Neurosci. 12, 13–31. (Accepted 9 May 1996)
PII:
Neuroscience Vol. 75, No. 2, pp. 339–344, 1996 Copyright ? 1996 IBRO. Published by Elsevier Science Ltd Printed in Great Britain 0306–4522/96 $15.00+0.00 S0306-4522(96)00301-6
Letter to Neuroscience TONIC FACILITATION OF GLUTAMATE RELEASE BY PRESYNAPTIC N-METHYL--ASPARTATE AUTORECEPTORS IN THE ENTORHINAL CORTEX N. BERRETTA* and R. S. G. JONES† Department of Pharmacology, University of Oxford, Mansfield Road, Oxford OX1 3QT, U.K. Key words: entorhinal cortex, NMDA, presynaptic receptors, glutamate release.
N-Methyl-D-aspartate receptors are fundamental for neuronal plasticity and development in the CNS.3,22 Most studies have examined postsynaptic roles of this receptor, but there are also indications for a presynaptic location and function.6,18,23 Here, we provide electrophysiological evidence for the existence of presynaptic N-methyl-D-aspartate receptors which can tonically facilitate glutamate release in the CNS. The N-methyl-D-aspartate receptor antagonist 2-amino-5phosphonopentanoate reduced the frequency, but not amplitude, of glutamate-mediated spontaneous excitatory postsynaptic currents in layer II neurons of the rat entorhinal cortex. This effect was also observed in the presence of tetrodotoxin and when postsynaptic N-methyl-D-aspartate receptors were blocked by dialysis with dizocilpine maleate. When extracellular calcium was replaced with strontium, 2-amino5-phosphonopentanoate reduced the ‘‘tail’’ of spontaneous excitatory postsynaptic currents that followed an evoked excitatory postsynaptic current. Finally, there was a tendency for paired-pulse facilitation of excitatory postsynaptic currents evoked at short (50 ms) intervals with postsynaptic N-methyl-Daspartate receptors blocked) to be reduced by 2-amino5-phosphonopentanoate, although this did not reach significance. These data strongly support the presence of presynaptic N-methyl-D-aspartate autoreceptors *Present address: SISSA—ISAS International School for Advanced Studies, Via Beirut 2/4, 34014, Trieste, Italy. †To whom correspondence should be addressed. Abbreviations: AQE, asynchronous quantal event; BAPTA, 1,2-bis(2-aminophenoxy)ethane-N,N,N*,N*-tetraacetic acid; -AP5, 2-amino-5-phosphonopentanoate; EC, entorhinal cortex; EGTA, ethyleneglycolbis(aminoethyl ether)tetra-acetate; EPSC, excitatory postsynaptic current; HEPES, N-2-hydroxyethylpiperazine-N*-2-ethanesulphonic acid; mEPSC, miniature EPSC; MK-801, dizocilpine maleate; NMDA, N-methyl--aspartate; PPF, paired-pulse facilitation; QX-314, N-(2,6-dimethyl phenylcarbamoylmethyl) triethylammonium bromide; sEPSC, spontaneous EPSC.
which may facilitate glutamate release in layer II of the entorhinal cortex. Copyright ? 1996 IBRO. Published by Elsevier Science Ltd. We have recently shown that neurons of the rat entorhinal cortex (EC) display continuous, spontaneous synaptic excitation mediated by release of glutamate, acting primarily at non-N-methyl-aspartate (NMDA) receptors.2 Application of the NMDA receptor antagonist 2-amino-5-phosphonopentanoate (-AP5; 100 µM, n = 6) did not affect the mean amplitude of spontaneous excitatory postsynaptic currents (sEPSCs) in these neurons (9.5 & 0.1 and 9.1 & 0.1 pA, before and after -AP5). Because the amplitudes of these events are not normally distributed,2 a non-parametric comparison of cumulative probability distributions was performed using the Kolmogorov–Smirnov test.24 This confirmed the lack of effect on amplitude of sEPSCs. However, a similar comparison of frequency of these events showed a clear reduction (P < 0.001, Kolmogorov–Smirnov test) in the presence of the antagonist (see Fig. 1). The mean interval between events increased significantly, from 135.1 & 2.7 to 194.8 & 5.4 ms, in the presence of -AP5. We have considered a number of interpretations of our observations. (1) -AP5 could cause a reduction of activity-dependent sEPSCs, normally driven by NMDA receptors in excitatory networks within the EC. (2) Postsynaptic NMDA receptors could induce an increase in presynaptic release via production of a retrograde messenger.4 (3) The frequency of sEPSCs could be set by a tonic feedback facilitatory action of glutamate on presynaptic NMDA receptors. To test the first two possibilities, we studied the effect of -AP5 on activity-independent, miniature excitatory postsynaptic currents (mEPSCs) recorded in the presence of tetrodotoxin (1 µM). To obviate a role of postsynaptic NMDA receptors, we included the open-channel NMDA receptor blocker,
339
340
N. Berretta and R. S. G. Jones
Fig. 1. -AP5 reduces the frequency but not amplitude of sEPSCs in layer II. (a) The traces in the upper panels are consecutive records of spontaneous activity recorded in a layer II neuron in the absence (left) and presence (right) of -AP5 (100 µM). The NMDA antagonist reduced the frequency of sEPSCs. The graphs show the cumulative distributions of amplitude and interval obtained by collecting 280 sEPSCs before (solid line) and after (dashed line) -AP5, from the same neuron. They show a significant (P < 0.001) increase in interval between events, with no change in amplitude. (b) Pooled data obtained by collecting 100 events from each of six neurons before (solid line) and after (dashed line) -AP5. Again, the interval but not the amplitude distribution was significantly (P < 0.001) altered by -AP5. Experiments were performed in slices (450 µm) of the EC, prepared from Wistar rats (120–150 g) as described previously.16 They were maintained at 34 & 0.2)C between a continuous (1.2 ml/min) perfusion with artificial cerebrospinal fluid and warmed, humidified carbogen gas. The artificial cerebrospinal fluid contained (in mM): NaCl 126, KCl 3, NaHCO3 24, NaH2PO4 1.25, MgSO4 2, CaCl2 2, -glucose 10. -AP5 (100 µM) was applied by bath perfusion or as a 5-µl droplet directly on to the slice. Whole-cell recordings were obtained using the ‘‘blind-patch’’ technique with glass electrodes filled with (in mM): CsMeSO4 130, HEPES 5, EGTA 0.5, MgCl2 1, NaCl 1, CaCl2 0.34, QX-314 5 (pH 7.3–7.4). Currents were recorded at a holding potential of "70 mV using an Axopatch 200A amplifier. They were filtered at 2 kHz, digitized at 48 kHz and recorded on a DAT recorder. Compensation for series resistance was not employed and statistical comparisons were only made when series resistance did not change by more than 10%. Spontaneous events were detected automatically using a threshold crossing criterion. The threshold level varied from neuron to neuron but was always the same before and after drug application. Peak amplitude and interval distributions were compared using the K-S test. Synaptic responses were also evoked by electrical stimulation via a bipolar platinum stimulating electrode placed on to the surface of the slice in the deep layers (V and VI) of the EC. Data are expressed as mean & S.E.M.
Glutamate release by NMDA autoreceptors in the EC
341
Fig. 2. Blockade of postsynaptic NMDA receptors does not affect the reduction in frequency of sEPCs induced by -AP5. (a) The records in the upper line show electrically evoked synaptic responses (stimulus at arrowheads) recorded in a layer II neuron extensively dialysed without MK-801 in the filling solution. The fast excitatory postsynaptic current (EPSC) was followed by a fast GABAA receptor-dependent inhibitory postsynaptic current. Following perfusion with 6-nitro-7-sulphoamoylbenzo(f)-quinoxaline-2,3dione (10 µM) and bicuculline methochloride (10 µM) a slow EPSC remained. This increased dramatically in magnesium-free medium and was abolished by -AP5 (100 µM), showing that it was mediated by NMDA receptors. The lower line shows a different neuron dialysed with a filling solution containing MK-801 (1 mM). An inhibitory postsynaptic current was not detectable in this neuron and, following perfusion with 6-nitro-7-sulphoamoylbenzo(f)-quinoxaline-2,3-dione plus bicuculline methochloride, there was clearly no slow EPSC, even when magnesium was removed from the perfusion medium. This neuron was also recorded at a holding potential of "40 mV to maximize any contribution from NMDA receptors. Thus, it seems highly probable that the postsynaptic NMDA receptors of the recorded neuron were blocked by internal application of MK-801. (b) The cumulative distributions of peak amplitude and intervals of sEPSC (n = 400) events before (solid line) and after (dashed line) perfusion with -AP5 (100 µM), from a neuron recorded in the presence of tetrodotoxin (1 µM) and bicuculline methochloride (10 µM) and with MK-801 in the patch pipette. -AP5 still significantly (P < 0.001) increased the interval between sEPSCs, without a change in amplitude. In these experiments the filling solution contained (in mM): CsMeSO4 122, CsCl 3, HEPES 5, BAPTA 10, (+)-MK-801 1, QX-314 5 (pH 7.3–7.4).
dizocilpine maleate (MK-801; 1 mM), in the recording pipette. Control studies clearly showed that when cells were dialysed with MK-801, the NMDA receptor-mediated component of electrically evoked synaptic responses was completely absent (n = 10; Fig. 2a). However, when -AP5 was applied in the presence of tetrodotoxin plus intracellular MK-801, it still caused a significant increase in the interval between mEPSCs, from 138 & 5 to 169 & 8 ms (P < 0.05, n = 9). Again, the amplitude of the same events was unaffected (11.1 & 0.5 and 11.3 &
0.6 pA, before and after -AP5; Fig. 2b). Thus, we believe that a likely explanation for the reduction in frequency by -AP5 is blockade of presynaptic NMDA receptors which tonically facilitate glutamate release. Substitution of extracellular calcium with strontium has been shown to facilitate ‘‘asynchronous’’ release of neurotransmitter. Thus, in the presence of strontium, stimulation of afferent axons elicits a reduced synaptic response succeeded by a series of asynchronous quantal events (AQEs), which result
342
N. Berretta and R. S. G. Jones
Fig. 3. -AP5 reduces the number of AQEs following an evoked response in the presence of strontium. In these experiments, CaCl2 was replaced with SrCl2 (2 mM) in the perfusion medium. MK-801 was included in the pipette solution. The traces in the upper panel show EPSCs evoked by electrical stimulation at 0.1 Hz. The point of stimulation is marked by the arrowheads. The five consecutive responses superimposed on the top line show that the initial evoked response was followed by a series of AQEs. The lower line shows that -AP5 (100 µM) had no effect on the initial evoked response, but reduced the occurrence of the succeeding AQEs. Results are quantified in Table 1. Table 1. Pooled data from six neurons recorded with strontium substituted for calcium, showing the amplitude (upper line) and number (lower line) of asynchronous quantal events detected in 40 successive traces within 300 ms of the peak of the evoked response
Amplitude (pA) Number
Control
-AP5
10.9 & 0.4 4.7 & 0.7
10.5 & 0.3 3.3 & 0.6
The number of events was significantly reduced in -AP5 (P < 0.02, t-test paired data), but their mean amplitude was unaltered.
from transmitter release from the recently activated synapses.12,21 Such an effect was seen in layer II neurons and permitted us to examine transmitter release occurring after stimulation of the afferent pathway. Taking into consideration the fact that NMDA receptors are permeable to strontium,19 we have addressed the possibility that presynaptic NMDA receptors might modulate this effect. Neurons in this study were again recorded with MK-801 in the patch pipettes. The number of sEPSCs recorded in a period of 300 ms following the peak of the evoked response was decreased significantly by -AP5 (100 µM; Fig. 3). However, the mean amplitude of the events that remained was unchanged.
These data are summarized in Table 1 and suggest that NMDA receptors could presynaptically facilitate release in the previously activated afferents, perhaps as a result of an increase in calcium influx into the terminals. Paired-pulse facilitation (PPF) is also considered to be a presynaptic phenomenon, resulting from a transient increase of presynaptic calcium, caused by a conditioning synaptic response.13,27 We have examined whether presynaptic NMDA receptors might contribute to PPF. PPF was evident 20–200 ms after the conditioning response in neurons (n = 6) recorded with MK-801 in the patch pipettes (Fig. 4). Perfusion with -AP5 reduced the degree of facilitation, particularly at the 50 ms interval (Fig. 4). Although the change was not statistically significant, the reduction of the facilitation by -AP5 was consistent in every cell and restricted to the 50 ms interval. It is quite likely that PPF would occur as a result of calcium entry via both presynaptic NMDA receptors and voltage-gated calcium channels, and that the contribution from the latter is greater for activitydependent responses. Thus, the present results are consistent with the presence of presynaptic NMDA autoreceptors on glutamate terminals in layer II of the EC. We cannot definitely rule out the possibility that the receptors are subsynaptic at axoaxonic glutamate synapses on
Glutamate release by NMDA autoreceptors in the EC
Fig. 4. -AP5 can reduce PPF in neurons where postsynaptic NMDA receptors were blocked. (a) The panel on the left shows the superimposed evoked EPSCs in a layer II neuron with inter-pulse intervals of 50 and 200 ms. The point of stimulation is marked by the arrowheads. Each trace is the average of five sweeps. The right-hand traces show that -AP5 (100 µM) reduced the degree of PPF at 50 ms but not at 200 ms. (b) Pooled data from six neurons show a trend towards a reduction in facilitation, particularly at 50 ms, although overall this effect just failed to reach significance. All experiments included MK-801 in the patch pipette filling solutions.
the glutamate terminals, but to our knowledge such synapses have not been demonstrated in the EC. NMDA autoreceptors appear to be activated by spontaneously released glutamate and determine, by positive feedback, a tonic facilitation of glutamate release. Binding sites for NMDA on presynaptic terminals have been demonstrated immunocytochemically.23 A facilitatory effect of NMDA on monoamine release has been shown,10,17,20,26 and on the release of acetylcholine at developing neuromuscular synapses.11 Neurochemical experiments have also indicated a presynaptic facilitation of glutamate release by NMDA receptors.6,18 However, electrophysiological evidence for presynaptic autoreceptors has previously been lacking. Autoreceptor modulation of glutamate release has largely been considered a domain of metabotropic receptors,1,8,25 although presynaptic kainate receptors have also been reported to be involved in control of glutamate
343
release.7,9 Our results show that NMDA receptors may be positively involved in such modulation as well. Presently, we can only speculate on the mechanism of the potentiation of release by presynaptic NMDA receptors. Perhaps the most likely explanation is that presynaptic NMDA receptors increase calcium influx via the receptor-gated ionophore. We need to explore this possibility. Another question which we have to ask is whether transmitter release at all glutamate synapses is modulated by presynaptic NMDA receptors, and at present we think this is unlikely to be the case. In layer V neurons, there is a greater contribution of postsynaptic NMDA receptors to the amplitude and duration of sEPSCs than in layer II,2 but we have found little effect of -AP5 on their frequency (Berretta and Jones, unpublished). In CA1 neurons, Oliet et al.21 have reported that long-term potentiation is associated with an increase in amplitude and frequency of AQEs recorded in strontium and the reverse effect occurs with long-term depression. Both effects are blocked by -AP5, but the authors do not indicate whether the antagonist had any effect on the frequency of AQEs per se. The functional role of presynaptic NMDA receptors is also purely speculative at present. Repetitive activation of afferents to layer II results in a shortterm, frequency-dependent enhancement of glutamate transmission,15 and it has been suggested that this may be involved in processing of information destined for the hippocampus.14 Presynaptic NMDA receptors could play a role in this short-term enhancement of transmission in the EC. Also, there is a strong school of thought which suggests that an increase in glutamate release is responsible, in part, for the phenomenon of long-term potentiation.4 Although this is suggested to occur subsequent to postsynaptic release of a retrograde messenger, it would not be unreasonable to suggest that increased release could also involve long-term presynaptic changes at an NMDA autoreceptor. Finally, it is noteworthy that layer II of the EC suffers very early and severe neurodegeneration in Alzheimer’s disease,5 and it could be speculated that this might involve an over-stimulation of glutamate release as a result of increased activity of presynaptic NMDA receptors.
Acknowledgements—We thank the Wellcome Trust and Royal Society for financial support, P. Vincent for the software for the analysis program and C. Stoub for the Kolmogorov–Smirnov test.
REFERENCES
1. 2.
Baskys A. and Malenka R. C. (1991) Agonists of metabotropic glutamate receptors presynaptically inhibit EPSCs in neonatal rat hippocampus. J. Physiol., Lond. 444, 687–701. Berretta N. and Jones R. S. G. (1996) A comparison of spontaneous EPSCs in layer II and layer IV–V neurones of the rat entorhinal cortex in vitro. J. Neurophysiol. 76, 1–12.
344 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27.
N. Berretta and R. S. G. Jones Blanton M. G., LoTurco J. J. and Kriegstein A. R. (1989) Whole-cell recording from neurones in slices of reptilian and mammalian cerebral cortex. J. Neurosci. Meth. 30, 203–210. Bliss T. V. P. and Collingridge G. L. (1993) A synaptic model of memory: long-term potentiation in the hippocampus. Nature 361, 31–39. Braak H. and Braak E. (1991) Neuropathological staging of Alzheimer-related changes. Acta neuropath. 82, 239–259. Bustos G., Abarca J., Forray M. I., Gysfing K., Bradberry C. W. and Roth R. H. (1992) Regulation of excitatory amino acid release by N-methyl--aspartate receptors in rat striatum: in vivo microdialysis studies. Brain Res. 585, 105–115. Chittajallu R., Vignes M., Dev K. K., Barnes J. M., Collingridge G. L. and Henley J. M. (1996) Regulation of glutamate release by presynaptic kainate receptors in the hippocampus. Nature 379, 78–81. Desai C. H., Smith T. S. and Conn P. J. (1992) Multiple metabotropic glutamate receptors regulate hippocampal function. Synapse 12, 206–213. Ferkany J. W., Zaczek R. and Coyle J. T. (1982) Kainic acid stimulates excitatory amino acid neurotransmitter release at presynaptic receptors. Nature 298, 757–759. Fink K., Bo¨nisch H. and Gothert M. (1990) Presynaptic NMDA receptors stimulate noradrenaline release in the cerebral cortex. Eur. J. Pharmac. 185, 115–117. Fu W.-M., Liou J.-C., Lee Y.-H. and Liou H.-C. (1995) Potentiation of neurotransmitter release by activation of presynaptic glutamate receptors at developing neuromuscular synapses of Xenopus. J. Physiol., Lond. 489, 813–823. Goda Y. and Stevens C. F. (1994) Two components of transmitter release at a central synapse. Proc. natn. Acad. Sci. U.S.A. 91, 12942–12946. Hess G., Kuhnt U. and Voronin L. L. (1987) Quantal analysis of paired-pulse facilitation in guinea pig hippocampal slices. Neurosci. Lett. 77, 187–192. Jones R. S. G. (1993) Entorhinal–hippocampal connections: a speculative view of their functions. Trends Neurosci. 16, 58–64. Jones R. S. G. (1994) Synaptic and intrinsic properties of neurones of origin of the perforant path in layer II of the rat entorhinal cortex in vitro. Hippocampus 4, 335–353. Jones R. S. G. and Heinemann U. (1988) Synaptic and intrinsic responses of medial entorhinal cortical cells in normal and magnesium-free medium in vitro. J. Neurophysiol. 59, 1476–1496. Krebs M. O., Desce J. M., Kemel M. L., Gauchy C., Godeheu G., Cheramy A. and Glowinski J. (1991) Glutamatergic control of dopamine release in the rat striatum: evidence for presynaptic N-methyl--aspartate receptors on dopaminergic nerve terminals. J. Neurochem. 56, 81–85. Martin D., Bustos G. A., Bowe M. A., Bray S. D. and Nadler J. V. (1991) Autoreceptor regulation of glutamate and aspartate release from slices of the hippocampal CA1 area. J. Neurochem. 56, 1647–1655. Mayer M. L. and Westbrook G. L. (1987) Permeation and block of N-methyl--aspartic acid receptor channels by divalent cations in mouse cultured central neurones. J. Physiol., Lond. 394, 501–527. Moghaddam B., Gruen R. J., Roth R. H., Bunney B. S. and Adams R. N. (1990) Effect of -glutamate on the release of striatal dopamine: in vivo dialysis and electrochemical studies. Brain Res. 518, 55–60. Oliet S. H. R., Malenka R. C. and Nicoll R. A. (1995) Bidirectional control of quantal size by synaptic activity in the hippocampus. Science 271, 1294–1297. Scheetz A. J. and Constantine-Paton M. (1994) The modulation of NMDA receptor function: implications for vertebrate neural development. Fedn Proc. Fedn Am. Socs exp. Biol. 8, 745–752. Siegel S. J., Brose N., Janssen W. G., Gasic G. P., Jahn R., Heinemann S. F. and Morrison J. H. (1994) Regional, cellular, and ultrastructural distribution of N-methyl--aspartate receptor subunit 1 in monkey hippocampus. Proc. natn. Acad. Sci. U.S.A. 91, 564–568. Van de Kloot W. (1991) The regulation of quantal size. Prog. Neurobiol. 36, 93–130. Vignes M., Clarke V. R. J., Davies C. H., Chambers A., Jane D. E., Watkins J. C. and Collingridge G. L. (1995) Pharmacological evidence for an involvement of group II and group III mGluRs in the presynaptic regulation of excitatory synaptic responses in the CA1 region of rat hippocampal slices. Neuropharmacology 34, 973–982. Wang J. K. T., Andrews H. and Thukral V. (1992) Presynaptic glutamate receptors regulate noradrenaline release from isolated nerve terminals. J. Neurochem. 58, 204–211. Zucker R. S. (1989) Short-term synaptic plasticity. A. Rev. Neurosci. 12, 13–31. (Accepted 9 May 1996)