- Email: [email protected]
NMDA receptor-dependent high-frequency network oscillations (100–300 Hz) in rat hippocampal slices
Neuroscience Letters 414 (2007) 197–202
NMDA receptor-dependent high-frequency network oscillations (100–300 Hz) in rat hippocampal slices Costas Papatheodoropoulos ∗ University of Patras, Department of Physiology, Medical School, 26504 Rio, Patras, Greece Received 26 July 2006; received in revised form 12 October 2006; accepted 24 October 2006
Abstract High-frequency oscillations (HFOs or ripples, ≥100 Hz) appear to be important expressions of cortical circuits, characterizing physiological and pathological functional states. Synaptic and non-synaptic mechanisms are involved in their generation. This study shows that spontaneous N-methyl-d-aspartate receptor (NMDAR) mediated potentials, recorded in dorsal and ventral hippocampal slices perfused with magnesium-free medium and antagonists of non-NMDARs and GABA receptors were associated with high-frequency oscillations (100–300 Hz), recorded in all hippocampal subregions. Both CA3 and CA1 regions displayed HFOs at the range of 180–300 Hz with oscillations in CA3 being significantly faster than in CA1 (232 ± 22 Hz, n = 64 slices versus 206 ± 18 Hz, n = 24, P < 0.001). Moreover, in most of the slices (39/63) the CA1 network oscillated also at a lower frequency (121.8 ± 2.45 Hz). Simultaneous recordings showed that activity was most often initiated in CA3 region; however, dentate gyrus and CA1 were potential sites of generation as well. The incidence of spontaneous events was significantly higher in ventral than in dorsal slices (20 ± 1.6/min versus 5.4 ± 0.3/min, P < 0.001). The competitive and non-competitive NMDAR antagonists, d-AP5 (50 M) and MK 801 (50 M), respectively abolished spontaneous activity. The gap-junction blocker carbenoxolone significantly suppressed spontaneous activity in a concentration-dependent manner. These data indicate that synaptic transmission provided by solely NMDARs can sustain the generation of high-frequency network oscillations, which display distinct characteristics in CA3 and CA1 subregions. © 2006 Elsevier Ireland Ltd. All rights reserved. Keywords: Ripples; High-frequency oscillations; NMDA; Excitatory potentials; Network synchronization; In vitro hippocampal slices
Hippocampal circuits exhibit characteristic high-frequency oscillations (HFOs ≥100 Hz) associated with certain physiological and pathological brain states, like slow wave sleep [6] and epileptogenesis [4,37]. Several in vitro models have shown that HFOs can be generated by mechanisms relying on chemical synaptic and electrical neuronal communications [11,14,26,37]. In models where chemical transmission is present, HFOs depend on the activation of non-NMDARs and/or GABAA receptors [11,23,30]. Still, the role of NMDARs is not clear, since in two studies no participation of these receptors has been found [11,23], whereas in two other studies the NMDAR blockage either enhanced [8] or reduced the amplitude of HFOs [30]. It seems therefore that the role of NMDARs in HFOs, if any, is merely modulatory. On the other hand, some examples of low frequency single neuron and network oscillations (at the range of ␦, and ␣ rhythms) appear to critically involve NMDARs
∗
Tel.: +30 2610 969117; fax: +30 2610 997215. E-mail address: [email protected].
0304-3940/$ – see front matter © 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.neulet.2006.10.036
[2,39]. Taking into consideration the important role of NMDARs in information processing (for review see [18]) and the growing appreciation of the role of HFOs in brain functions [4,37] it is questioned whether excitatory synaptic transmission based on NMDARs can sustain HFOs of the hippocampal network. Here, it is shown that hippocampal HFOs can be based on synaptic transmission mediated by only NMDARs. Transverse dorsal and ventral hippocampal slices were prepared from 20, 2 months old, male Wistar rats as previously described [27]. All efforts were made to minimize the number and suffering of animals, according to the European Communities Council Directive Guidelines (86/609/EEC) for the care and use of Laboratory animals. Animals, after deep anesthesia with diethyl-ether, were decapitated, the brain was placed in chilled (2–4 ◦ C) standard medium (containing, in mM: 124 NaCl; 4 KCl; 2 MgSO4 ; 2 CaCl2 ; 1.25 NaH2 PO4 ; 26 NaHCO3; 10 glucose; at pH 7.4) and the two hippocampi were excised free. Using a chopper, 550 m thick slices were prepared and immediately transferred to an interface type chamber continuously humidified with gas 95% O2 /5% CO2 , perfused with standard medium at
198
C. Papatheodoropoulos / Neuroscience Letters 414 (2007) 197–202
32 ± 0.5 ◦ C. After about 30 min magnesium ions were removed from the perfusate. Furthermore, in order to isolate NMDARmediated potentials the antagonists of ionotropic non-NMDARs 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX, 20–40 M), of GABAA receptors picrotoxin (50 M), and of GABAB receptors (3-aminopropyl)(diethoxymethyl)phosphinic acid (CGP 35348, 75–100 M) were added to the medium. All drugs were purchased from Tocris (UK), but picrotoxin was obtained from Sigma (USA). Field potentials were recorded from the somatic or dendritic layer of the subregions using carbon-fiber electrodes (diameter 10 m, WPI, USA). A bipolar platinum/iridium electrode (25 m) was used for electrical stimulation, every 30 s. Records were filtered at 0.5–2 kHz, digitized at 4 kHz and stored on a computer disk using the CED 1401-plus system and the Spike and Signal software (Cambridge Electronic Design, UK). In spontaneous recordings, the dominant high-frequency peaks were determined by power spectra computed in 3–5 min lasting epochs of raw data. Using band pass (80–550 Hz) filtered records, measures of activity delays between subregions were performed by detecting the start point of oscillation as the sweep deflection greater than four times the standard deviation of the event-free baseline noise [23]. At the end of some experiments, after washing out the drugs (for at least 3 h) and reintroducing magnesium ions in the medium, the evoked responses were recorded in order to examine the slice for health condition. In all 27 slices tested (obtained from three animals), the field potentials were normal (Fig. 1G). Paired and independent t-tests and the bivariate two-tailed correlation test were used for statistical comparisons. Values in the text are given as mean ± S.E.M. and numbers inside parenthesis indicate the number of slices studied. Spontaneous NMDAR-mediated field potentials were studied in 57 dorsal and 45 ventral slices. Recordings of spontaneous activity started at about 1 h after introducing the drug cocktail in the medium. In stratum pyramidale, granule cell layer of dentate gyrus (DG) and dentate hilus spontaneous events consisted of sharp wave-like potentials ridden by burst-like activity (Fig. 1A and B). Power spectra analysis showed that the latter had a dominant high-frequency peak in the range of 100–300 Hz (inserts in Fig. 1A and B), similar to that previously observed in the rodent hippocampus under various experimental conditions. Spontaneous potentials and the associated HFOs recurred at a rate of 2–49 events/min. However, although the minimum rate values were similar between dorsal (2/min) and ventral (3/min) slices, the range of rate values was broader in ventral slices, which showed significantly increasing rates as they were taken from more extreme positions (Fig. 1C). Thus, ventral slices displayed a significantly higher rate than dorsal ones (20 ± 1.6 versus 5.4 ± 0.3, P < 0.001). HFOs were studied mainly in CA3 and CA1, where they were more consistently observed. Recordings from CA3 and CA1 were made from 71 and 63 slices, respectively. Simultaneous measures from both subregions were taken from 42 slices. The histograms of peak frequency values in CA3 and CA1 showed that in the great majority of slices (64/71, 90%) HFO peak values in CA3 were found only at a relatively highfrequency range (180–300 Hz), (Fig. 1D and E). On the other hand, peak frequency values in the CA1 region of most slices (39/63, 62%) fell in a second range of slower frequencies, from
100 to 170 Hz (121.78 ± 15/2.45 Hz) while in the rest of cases (24/63, 38%) activity in CA1 displayed a dominant frequency at the higher-frequency range (>180 Hz), similar to CA3 peaks (Fig. 1D and E). Oscillations at the higher-frequency range were significantly faster in CA3 (231.9 ± 22.4/2.8 Hz, n = 64 slices) than in CA1 (205.75 ± 3.7 Hz, n = 24 slices). This was observed also within individual slices (see Fig. 2A). The frequency did not differ between dorsal and ventral slices, either in CA3 (218.7 ± 6.3 Hz, n = 38 dorsal slices versus 227.35 ± 5.6 Hz, n = 33 ventral slices) or in CA1 region (157.8 ± 7.3 Hz, n = 38 dorsal slices versus 149.4 ± 8.6 Hz, n = 25 ventral slices). However, in CA1, while similar numbers of dorsal slices oscillated at the two frequency bands (21/38 slices with 100–150 Hz and 17/38 slices with >180 Hz), ventral slices displayed most often frequencies preferentially at the lower-frequency range (17/25, 68%, Fig. 1F). In line with previous studies [10,11,23], HFOs most often initiated in the CA3 region and spread toward CA1 (34 out of 37 slices, Fig. 2A), with a mean delay of 8.21 ± 1.19 ms (n = 10). In three experiments, made in two ventral and one dorsal slice, activity was initiated in the CA1 region or it was not be possible to detect any delay between the two recordings (Fig. 2B and C). Interestingly, in two of those experiments and in another, in which activity in CA3 led that in CA1, HFOs in CA1 was occasionally independent from that in CA3, (Fig. 2D and E). In 14 slices, simultaneous recordings were made from CA3 and DG or hilus. In 11 of those, experiments activity was reliably initiated in CA3 and propagated into DG with a delay of 15.41 ± 1.2 ms (n = 6), while in the other three cases DG led CA3 (delay 9.21 ± 3.18 ms, n = 3). In one slice, activity which was initiated in CA3 occasionally failed to propagate to the DG. Spontaneous activity was abolished after application of either the competitive or non-competitive antagonists of NMDARs, DAP5 (50 M, n = 6) and MK-801 (50 M, n = 10), respectively, (Fig. 3). The time required to suppress activity was 6.5 min for D-AP5 and 17.8 min for MK-801. The activity suppression was reversed following >40 min drug wash-out (n = 2). Electrical coupling through gap-junctions underlies generation of HFOs in the absence of chemical synaptic transmission [14], and the role of gap-junctions in HFOs has been pointed out using several experimental models [37], but some studies have questioned this relationship [11,30]. In this study, the possible involvement of gap-junctions on HFOs was examined by perfusing the slices with the gapjunction blocker carbenoxolone. The drug, at the concentrations of 0.1 mM produced a significant decrease in the rate of spontaneous events, from 17.1 ± 4.3 to 6.3 ± 1.4 events/min (n = 11, P < 0.05); however, it did not consistently affect the frequency (188.6 ± 16.2 Hz versus 185.6 ± 10 Hz, in control and drug conditions, respectively. Fig. 3C and E). At the concentrations of 0.2 mM carbenoxolone eventually abolished spontaneous activity (the rate was decreased from 17.85 ± 4.91 to 1.13 ± 0.46 events/min, n = 10, P < 0.0001, Fig. 2D and E). Carbenoxolone did not affect evoked NMDAR-mediated synaptic potentials (three slices, Fig. 3F). This study demonstrates that HFOs in the hippocampus can be sustained by synaptic transmission mediated solely by
C. Papatheodoropoulos / Neuroscience Letters 414 (2007) 197–202
199
Fig. 1. HFOs in the CA3 and CA1 regions of dorsal and ventral hippocampal slices. (A) Examples of field recordings from the CA3 st. pyramidale of a dorsal and a ventral slice (upper and lower panel, respectively). Continuous recordings of spontaneous activity (traces on the left) and single events (traces on the right) are shown. Note the lower rate of spontaneous events in the dorsal slice. Power spectra (inserts) illustrate the high-frequency peaks (200–250 Hz) revealed after fast fourier transformation. (B) Single events (traces on the left) recorded from DG and hilus of two different ventral slices. The corresponding power spectra are shown on the right. (C) Diagram where the rate of spontaneous events is plotted against the position of the slices along the dorsal–ventral axis of hippocampus. Note that the rate increases in ventral slices taken from progressively more extreme positions (positions 14–18, r = −0.73, P < 0.001) whereas the rate is similar among dorsal slices (positions 3–8). (D) Power spectra of CA3 (upper panel) and CA1 (lower panel) st. pyramidale recordings illustrating the higher frequency of activity in the CA3 compared to CA1 region. Inserts show wide-band (upper traces) and band pass (lower traces) recordings of single events. (E) Histograms of the high-frequency peaks observed in recordings from the CA3 and CA1 region. Note that most of the CA3 values were distributed at the range of 180–300 Hz, whereas CA1 presented an additional distribution at the range of 100–150 Hz. (F) Histograms of the CA3 high-frequency peaks in dorsal and ventral slices. (G) Example of CA1 evoked population spikes recorded from a slice in standard medium at the beginning of the experiment, i.e. before application of the drug cocktail (a) and at the end of the experiment, i.e. after three hours of washing out all drugs and reintroducing magnesium in the medium (b). Averages of four sweeps are shown. Stimulation artifacts are truncated.
NMDARs and that CA3 and CA1 display distinct features of such activity. Similarly to previous studies [5,11,26] HFOs were observed in the absence of fast GABAergic transmission. HFOs were observed in the CA3 and CA1 subregions in contrast to a recent study, in disinhibited slices, showing HFOs in the DG but not in CA3 and CA1, where paroxysmal events were seen [11].
This may be accounted for by the fact that excitation resting on non-NMDARs under complete blockade of inhibition [11] in combination with the large amount of recurrent excitatory collaterals in CA3 region [1] leads to exaggerated depolarizations which presumably disturb the generation of HFOs. In accord with this idea, HFOs have been observed in the disinhibited
200
C. Papatheodoropoulos / Neuroscience Letters 414 (2007) 197–202
Fig. 2. Both, CA3 and CA1 were sites of initiation of NMDA-dependent HFOs. (A) Simultaneous recordings (traces on the left column) of a ventral slice and corresponding power spectra (on the right), exemplifying the most frequent situation where oscillation was initiated in CA3 and propagated toward CA1 region. Dotted line is positioned at the start of CA3 activity. (B) An example of simultaneous recordings from CA3 and CA1 from a dorsal slice showing very small activity delay (∼2.5 ms) between the two regions. In this example, both regions displayed similar oscillation frequency. Note the large cross correlation of the two recordings (diagram on the right). (C) Representative example of the experimental cases where oscillation was initiated in the CA1 region. Recordings were obtained from a ventral slice. (D) Another example with activity initiation in CA1 region (a). On the same slice, activity was occasionally recorded in CA1 with only a very small signal in CA3 region. Data acquired from a dorsal slice. (E) An instance from simultaneous CA3 and CA1 recordings from another dorsal slice showing complete absence of activity in CA3 despite strong HFO in CA1 region. All recordings in the figure are band-pass filtered (80–550 Hz). However, all fast fourier transforms were obtained from unfiltered raw data.
CA1 minislice [26], in which the massive input from CA3 to CA1 was eliminated. Action potential bursts mediated by recurrent connections and intrinsic firing mechanisms are importantly involved in excitation-dependent HFOs [15,21]. Therefore, the recurrent connectivity and intrinsic firing properties, which are both different in CA3 and CA1 neurons [13], might explain to some extent the distinct oscillation frequencies seen in CA3 and CA1. The CA3 to DG propagation of activity, also previously shown [11,35], presumably resulted from the functionally uncovered, under disinhibition, CA3 to DG polysynaptic connections [35,36] with the crucial role of hilar neurons [7,35]. The independence of CA1 in generating HFOs, also previously observed [10,23,26,28], could be explained by the relatively higher content of NMDARs in CA1 [24] in combination with their involvement in CA1 polysynaptic activation under similar
to the present study conditions [9]. NMDARs in recurrent collaterals may provide a key element in recall memory mechanisms served by network oscillations [20]. Furthermore, NMDARs are strictly involved in memory processes carried out by specific hippocampal subregions [25,31]. Thus, the distinct frequency preference in the NMDAR-dependent HFOs displayed by the CA3 and CA1 hippocampal subregions in combination with the distinct NMDAR-dependent roles in memory processes taken by these subregions [22] corroborate the concept that information flow through hippocampus is subject to distinct processing by the discrete circuits of its subregions [21,32,38]. In line with previous reports [14,17,23,28], it is shown here that gap-junction blockade suppresses the generation of HFOs. Other studies however, have previously shown that gap-junction uncouplers failed to suppress HFOs induced either by disinhibition [11] or by tetanic stimulation [30]. It seems therefore
C. Papatheodoropoulos / Neuroscience Letters 414 (2007) 197–202
201
Fig. 3. Pharmacological examination of HFOs. (A and B) The antagonists of NMDARs D-AP5 and MK-801 consistently abolished spontaneous activity. Continuous recordings of spontaneous activity from CA3 (A) and CA1 (B) and the corresponding power spectra before (a) and during (b) drug application (black bars) are shown. Note that both drugs completely abolished spontaneous activity and the associated HFO in both CA3 and CA1 regions. (C and F) Effects of carbenoxolone on HFOs. (C and D) Continuous recordings of CA3 spontaneous activity (traces on the left) and the corresponding power spectra (diagrams on the right) from two slices. (C) The drug at the concentrations of 0.1 mM decreased the incidence of spontaneous potentials without significant effect on HFO frequency. At 0.2 mM (D) carbenoxolone strongly suppressed spontaneous potentials and accompanying HFOs (E (a), lower traces) but did not affect the evoked NMDAR-mediated synaptic response (E (b)) recorded from the same slice. F. Cumulative results showing the concentration-dependent effect of carbenoxolone on the rate of spontaneous activity.
that although electrical coupling is crucial in several models of high-frequency activity, in other models alternative mechanisms may contribute in synchronizing neuronal populations into coherent activity. It is also possible that the effects of carbenoxolone are mediated by a mechanism not involving gap junctions [33], although this seems unlikely given that the drug did not affect NMDAR-mediated responses. The higher incidence of spontaneous activity in ventral compared with dorsal slices is in line with previous data showing a positive correlation between neuronal excitability and the longitudinal axis of the hippocampus in the dorso-ventral direction [3,12,16,29]. The ventral compared to dorsal hippocampus
exhibits a higher probability of transmitter release at its excitatory synapses [27] and higher levels of extracellular glutamate [19]. These may contribute to the higher excitability of the ventral hippocampus observed in the present study, expressed as a higher incidence of spontaneous events, taking into account that extracellular glutamate enhances neuronal excitability by tonically activating NMDARs [34]. In conclusion, the present findings reveal NMDARs as an alternative source of excitatory drive in the generation of network HFOs. It seems that hippocampal circuits can achieve a high-frequency oscillatory state by engaging several distinct mechanisms. Furthermore, the oscillations in CA3 and CA1
202
C. Papatheodoropoulos / Neuroscience Letters 414 (2007) 197–202
hippocampal subregions expressed distinct characteristics suggesting that the microcircuitry of each region produces and sustains activity of this type in a distinct way. Acknowledgments Study funded by the European Social Fund (ESF), Operational Program for Educational and Vocational Training II (EPEAEK II), and particularly the Program PYTHAGORAS II. I wish to thank Dr. George Kostopoulos for helpful comments on the manuscript. References [1] D.G. Amaral, N. Ishizuka, B. Claiborne, Neurons, numbers and the hippocampal network, Prog. Brain Res. 83 (1990) 1–11. [2] C. Bonansco, W. Buno, Cellular mechanisms underlying the rhythmic bursts induced by NMDA microiontophoresis at the apical dendrites of CA1 pyramidal neurons, Hippocampus 13 (2003) 150–163. [3] C. Borck, J.G. Jefferys, Seizure-like events in disinhibited ventral slices of adult rat hippocampus, J. Neurophysiol. 82 (1999) 2130–2142. [4] A. Bragin, J. Engel Jr., C.L. Wilson, I. Fried, G. Buzsaki, High-frequency oscillations in human brain, Hippocampus 9 (1999) 137–142. [5] A. Bragin, I. Mody, C.L. Wilson, J. Engel Jr., Local generation of fast ripples in epileptic brain, J. Neurosci. 22 (2002) 2012–2021. [6] G. Buzsaki, Z. Horvath, R. Urioste, J. Hetke, K. Wise, High-frequency network oscillation in the hippocampus, Science 256 (1992) 1025–1027. [7] G.C. Carlson, Role of hilar neurons in the propagation of fast-ripples from CA3 into the dentate gyrus: a fast voltage-sensitive dye imaging study, Soc. Neurosci. Abstr. (2004) (No. 853.2.). [8] L.L. Colgin, Y. Jia, J.M. Sabatier, G. Lynch, Blockade of NMDA receptors enhances spontaneous sharp waves in rat hippocampal slices, Neurosci. Lett. 385 (2005) 46–51. [9] V. Crepel, R. Khazipov, Y. Ben-Ari, Blocking GABA(A) inhibition reveals AMPA- and NMDA-receptor-mediated polysynaptic responses in the CA1 region of the rat hippocampus, J. Neurophysiol. 77 (1997) 2071–2082. [10] J. Csicsvari, H. Hirase, A. Czurko, A. Mamiya, G. Buzsaki, Fast network oscillations in the hippocampal CA1 region of the behaving rat, J. Neurosci. 19 (1999) RC20. [11] M. D’Antuono, P. de Guzman, T. Kano, M. Avoli, Ripple activity in the dentate gyrus of dishinibited hippocampus-entorhinal cortex slices, J. Neurosci. Res. 80 (2005) 92–103. [12] M. Derchansky, E. Shahar, R.A. Wennberg, M. Samoilova, S.S. Jahromi, P.A. Abdelmalik, L. Zhang, P.L. Carlen, Model of frequent, recurrent, and spontaneous seizures in the intact mouse hippocampus, Hippocampus 14 (2004) 935–947. [13] J. Deuchars, A.M. Thomson, CA1 pyramid-pyramid connections in rat hippocampus in vitro: dual intracellular recordings with biocytin filling, Neuroscience 74 (1996) 1009–1018. [14] A. Draguhn, R.D. Traub, D. Schmitz, J.G. Jefferys, Electrical coupling underlies high-frequency oscillations in the hippocampus in vitro, Nature 394 (1998) 189–192. [15] V.I. Dzhala, K.J. Staley, Mechanisms of fast ripples in the hippocampus, J. Neurosci. 24 (2004) 8896–8906. [16] M. Gilbert, R.J. Racine, G.K. Smith, Epileptiform burst responses in ventral vs dorsal hippocampal slices, Brain Res. 361 (1985) 389–391. [17] S.J. Gladwell, J.G. Jefferys, Second messenger modulation of electrotonic coupling between region CA3 pyramidal cell axons in the rat hippocampus, Neurosci. Lett. 300 (2001) 1–4. [18] M. Hausser, N. Spruston, G.J. Stuart, Diversity and dynamics of dendritic signaling, Science 290 (2000) 739–744.
[19] H. Hortnagl, M.L. Berger, G. Sperk, C. Pifl, Regional heterogeneity in the distribution of neurotransmitter markers in the rat hippocampus, Neuroscience 45 (1991) 261–272. [20] O. Jensen, J.E. Lisman, Theta/gamma networks with slow NMDA channels learn sequences and encode episodic memory: role of NMDA channels in recall, Learn. Mem. 3 (1996) 264–278. [21] R.P. Kesner, I. Lee, P. Gilbert, A behavioral assessment of hippocampal function based on a subregional analysis, Rev. Neurosci. 15 (2004) 333–351. [22] I. Lee, R.P. Kesner, Differential contribution of NMDA receptors in hippocampal subregions to spatial working memory, Nat. Neurosci. 5 (2002) 162–168. [23] N. Maier, V. Nimmrich, A. Draguhn, Cellular and network mechanisms underlying spontaneous sharp wave-ripple complexes in mouse hippocampal slices, J. Physiol. 550 (2003) 873–887. [24] D.T. Monaghan, D. Yao, H.J. Olverman, J.C. Watkins, C.W. Cotman, Autoradiography of d-2-[3H]amino-5-phosphonopentanoate binding sites in rat brain, Neurosci. Lett. 52 (1984) 253–258. [25] K. Nakazawa, M.C. Quirk, R.A. Chitwood, M. Watanabe, M.F. Yeckel, L.D. Sun, A. Kato, C.A. Carr, D. Johnston, M.A. Wilson, S. Tonegawa, Requirement for hippocampal CA3 NMDA receptors in associative memory recall, Science 297 (2002) 211–218. [26] V. Nimmrich, N. Maier, D. Schmitz, A. Draguhn, Induced sharp wave-ripple complexes in the absence of synaptic inhibition in mouse hippocampal slices, J. Physiol. 563 (2005) 663–670. [27] C. Papatheodoropoulos, G. Kostopoulos, Dorsal–ventral differentiation of short-term synaptic plasticity in rat CA1 hippocampal region, Neurosci. Lett. 286 (2000) 57–60. [28] C. Papatheodoropoulos, G. Kostopoulos, Mechanisms of generation of a novel rhythmical network activity in ventral hippocampal slices, Soc. Neurosci. Abstr. (2003) (No. 582.8.). [29] C. Papatheodoropoulos, C. Moschovos, G. Kostopoulos, Greater contribution of N-methyl-d-aspartic acid receptors in ventral compared to dorsal hippocampal slices in the expression and long-term maintenance of epileptiform activity, Neuroscience 135 (2005) 765–779. [30] B. Poschel, U. Heinemann, A. Draguhn, High frequency oscillations in the dentate gyrus of rat hippocampal slices induced by tetanic stimulation, Brain Res. 959 (2003) 320–327. [31] T. Rajji, D. Chapman, H. Eichenbaum, R. Greene, The role of CA3 hippocampal NMDA receptors in paired associate learning, J. Neurosci. 26 (2006) 908–915. [32] R.G. Robertson, E.T. Rolls, P. Georges-Franois, Spatial view cells in the primate hippocampus: effects of removal of view details, J. Neurophysiol. 79 (1998) 1145–1156. [33] N. Rouach, M. Segal, A. Koulakoff, C. Giaume, E. Avignone, Carbenoxolone blockade of neuronal network activity in culture is not mediated by an action on gap junctions, J. Physiol. 553 (2003) 729–745. [34] P. Sah, S. Hestrin, R.A. Nicoll, Tonic activation of NMDA receptors by ambient glutamate enhances excitability of neurons, Science 246 (1989) 815–818. [35] H.E. Scharfman, Evidence from simultaneous intracellular recordings in rat hippocampal slices that area CA3 pyramidal cells innervate dentate hilar mossy cells, J. Neurophysiol. 72 (1994) 2167–2180. [36] H.E. Scharfman, Conditions required for polysynaptic excitation of dentate granule cells by area CA3 pyramidal cells in rat hippocampal slices, Neuroscience 72 (1996) 655–658. [37] R.D. Traub, Fast oscillations and epilepsy, Epilepsy Curr. 3 (2003) 77–79. [38] A. Treves, Computational constraints between retrieving the past and predicting the future, and the CA3–CA1 differentiation, Hippocampus 14 (2004) 539–556. [39] H. Yoshimura, N. Kato, M. Honjo, T. Sugai, N. Segami, N. Onoda, Agedependent emergence of a parieto-insular corticocortical signal flow in developing rats, Brain Res. Dev. Brain Res. 149 (2004) 45–51.
NMDA receptor-dependent high-frequency network oscillations (100–300 Hz) in rat hippocampal slices Costas Papatheodoropoulos ∗ University of Patras, Department of Physiology, Medical School, 26504 Rio, Patras, Greece Received 26 July 2006; received in revised form 12 October 2006; accepted 24 October 2006
Abstract High-frequency oscillations (HFOs or ripples, ≥100 Hz) appear to be important expressions of cortical circuits, characterizing physiological and pathological functional states. Synaptic and non-synaptic mechanisms are involved in their generation. This study shows that spontaneous N-methyl-d-aspartate receptor (NMDAR) mediated potentials, recorded in dorsal and ventral hippocampal slices perfused with magnesium-free medium and antagonists of non-NMDARs and GABA receptors were associated with high-frequency oscillations (100–300 Hz), recorded in all hippocampal subregions. Both CA3 and CA1 regions displayed HFOs at the range of 180–300 Hz with oscillations in CA3 being significantly faster than in CA1 (232 ± 22 Hz, n = 64 slices versus 206 ± 18 Hz, n = 24, P < 0.001). Moreover, in most of the slices (39/63) the CA1 network oscillated also at a lower frequency (121.8 ± 2.45 Hz). Simultaneous recordings showed that activity was most often initiated in CA3 region; however, dentate gyrus and CA1 were potential sites of generation as well. The incidence of spontaneous events was significantly higher in ventral than in dorsal slices (20 ± 1.6/min versus 5.4 ± 0.3/min, P < 0.001). The competitive and non-competitive NMDAR antagonists, d-AP5 (50 M) and MK 801 (50 M), respectively abolished spontaneous activity. The gap-junction blocker carbenoxolone significantly suppressed spontaneous activity in a concentration-dependent manner. These data indicate that synaptic transmission provided by solely NMDARs can sustain the generation of high-frequency network oscillations, which display distinct characteristics in CA3 and CA1 subregions. © 2006 Elsevier Ireland Ltd. All rights reserved. Keywords: Ripples; High-frequency oscillations; NMDA; Excitatory potentials; Network synchronization; In vitro hippocampal slices
Hippocampal circuits exhibit characteristic high-frequency oscillations (HFOs ≥100 Hz) associated with certain physiological and pathological brain states, like slow wave sleep [6] and epileptogenesis [4,37]. Several in vitro models have shown that HFOs can be generated by mechanisms relying on chemical synaptic and electrical neuronal communications [11,14,26,37]. In models where chemical transmission is present, HFOs depend on the activation of non-NMDARs and/or GABAA receptors [11,23,30]. Still, the role of NMDARs is not clear, since in two studies no participation of these receptors has been found [11,23], whereas in two other studies the NMDAR blockage either enhanced [8] or reduced the amplitude of HFOs [30]. It seems therefore that the role of NMDARs in HFOs, if any, is merely modulatory. On the other hand, some examples of low frequency single neuron and network oscillations (at the range of ␦, and ␣ rhythms) appear to critically involve NMDARs
∗
Tel.: +30 2610 969117; fax: +30 2610 997215. E-mail address: [email protected].
0304-3940/$ – see front matter © 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.neulet.2006.10.036
[2,39]. Taking into consideration the important role of NMDARs in information processing (for review see [18]) and the growing appreciation of the role of HFOs in brain functions [4,37] it is questioned whether excitatory synaptic transmission based on NMDARs can sustain HFOs of the hippocampal network. Here, it is shown that hippocampal HFOs can be based on synaptic transmission mediated by only NMDARs. Transverse dorsal and ventral hippocampal slices were prepared from 20, 2 months old, male Wistar rats as previously described [27]. All efforts were made to minimize the number and suffering of animals, according to the European Communities Council Directive Guidelines (86/609/EEC) for the care and use of Laboratory animals. Animals, after deep anesthesia with diethyl-ether, were decapitated, the brain was placed in chilled (2–4 ◦ C) standard medium (containing, in mM: 124 NaCl; 4 KCl; 2 MgSO4 ; 2 CaCl2 ; 1.25 NaH2 PO4 ; 26 NaHCO3; 10 glucose; at pH 7.4) and the two hippocampi were excised free. Using a chopper, 550 m thick slices were prepared and immediately transferred to an interface type chamber continuously humidified with gas 95% O2 /5% CO2 , perfused with standard medium at
198
C. Papatheodoropoulos / Neuroscience Letters 414 (2007) 197–202
32 ± 0.5 ◦ C. After about 30 min magnesium ions were removed from the perfusate. Furthermore, in order to isolate NMDARmediated potentials the antagonists of ionotropic non-NMDARs 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX, 20–40 M), of GABAA receptors picrotoxin (50 M), and of GABAB receptors (3-aminopropyl)(diethoxymethyl)phosphinic acid (CGP 35348, 75–100 M) were added to the medium. All drugs were purchased from Tocris (UK), but picrotoxin was obtained from Sigma (USA). Field potentials were recorded from the somatic or dendritic layer of the subregions using carbon-fiber electrodes (diameter 10 m, WPI, USA). A bipolar platinum/iridium electrode (25 m) was used for electrical stimulation, every 30 s. Records were filtered at 0.5–2 kHz, digitized at 4 kHz and stored on a computer disk using the CED 1401-plus system and the Spike and Signal software (Cambridge Electronic Design, UK). In spontaneous recordings, the dominant high-frequency peaks were determined by power spectra computed in 3–5 min lasting epochs of raw data. Using band pass (80–550 Hz) filtered records, measures of activity delays between subregions were performed by detecting the start point of oscillation as the sweep deflection greater than four times the standard deviation of the event-free baseline noise [23]. At the end of some experiments, after washing out the drugs (for at least 3 h) and reintroducing magnesium ions in the medium, the evoked responses were recorded in order to examine the slice for health condition. In all 27 slices tested (obtained from three animals), the field potentials were normal (Fig. 1G). Paired and independent t-tests and the bivariate two-tailed correlation test were used for statistical comparisons. Values in the text are given as mean ± S.E.M. and numbers inside parenthesis indicate the number of slices studied. Spontaneous NMDAR-mediated field potentials were studied in 57 dorsal and 45 ventral slices. Recordings of spontaneous activity started at about 1 h after introducing the drug cocktail in the medium. In stratum pyramidale, granule cell layer of dentate gyrus (DG) and dentate hilus spontaneous events consisted of sharp wave-like potentials ridden by burst-like activity (Fig. 1A and B). Power spectra analysis showed that the latter had a dominant high-frequency peak in the range of 100–300 Hz (inserts in Fig. 1A and B), similar to that previously observed in the rodent hippocampus under various experimental conditions. Spontaneous potentials and the associated HFOs recurred at a rate of 2–49 events/min. However, although the minimum rate values were similar between dorsal (2/min) and ventral (3/min) slices, the range of rate values was broader in ventral slices, which showed significantly increasing rates as they were taken from more extreme positions (Fig. 1C). Thus, ventral slices displayed a significantly higher rate than dorsal ones (20 ± 1.6 versus 5.4 ± 0.3, P < 0.001). HFOs were studied mainly in CA3 and CA1, where they were more consistently observed. Recordings from CA3 and CA1 were made from 71 and 63 slices, respectively. Simultaneous measures from both subregions were taken from 42 slices. The histograms of peak frequency values in CA3 and CA1 showed that in the great majority of slices (64/71, 90%) HFO peak values in CA3 were found only at a relatively highfrequency range (180–300 Hz), (Fig. 1D and E). On the other hand, peak frequency values in the CA1 region of most slices (39/63, 62%) fell in a second range of slower frequencies, from
100 to 170 Hz (121.78 ± 15/2.45 Hz) while in the rest of cases (24/63, 38%) activity in CA1 displayed a dominant frequency at the higher-frequency range (>180 Hz), similar to CA3 peaks (Fig. 1D and E). Oscillations at the higher-frequency range were significantly faster in CA3 (231.9 ± 22.4/2.8 Hz, n = 64 slices) than in CA1 (205.75 ± 3.7 Hz, n = 24 slices). This was observed also within individual slices (see Fig. 2A). The frequency did not differ between dorsal and ventral slices, either in CA3 (218.7 ± 6.3 Hz, n = 38 dorsal slices versus 227.35 ± 5.6 Hz, n = 33 ventral slices) or in CA1 region (157.8 ± 7.3 Hz, n = 38 dorsal slices versus 149.4 ± 8.6 Hz, n = 25 ventral slices). However, in CA1, while similar numbers of dorsal slices oscillated at the two frequency bands (21/38 slices with 100–150 Hz and 17/38 slices with >180 Hz), ventral slices displayed most often frequencies preferentially at the lower-frequency range (17/25, 68%, Fig. 1F). In line with previous studies [10,11,23], HFOs most often initiated in the CA3 region and spread toward CA1 (34 out of 37 slices, Fig. 2A), with a mean delay of 8.21 ± 1.19 ms (n = 10). In three experiments, made in two ventral and one dorsal slice, activity was initiated in the CA1 region or it was not be possible to detect any delay between the two recordings (Fig. 2B and C). Interestingly, in two of those experiments and in another, in which activity in CA3 led that in CA1, HFOs in CA1 was occasionally independent from that in CA3, (Fig. 2D and E). In 14 slices, simultaneous recordings were made from CA3 and DG or hilus. In 11 of those, experiments activity was reliably initiated in CA3 and propagated into DG with a delay of 15.41 ± 1.2 ms (n = 6), while in the other three cases DG led CA3 (delay 9.21 ± 3.18 ms, n = 3). In one slice, activity which was initiated in CA3 occasionally failed to propagate to the DG. Spontaneous activity was abolished after application of either the competitive or non-competitive antagonists of NMDARs, DAP5 (50 M, n = 6) and MK-801 (50 M, n = 10), respectively, (Fig. 3). The time required to suppress activity was 6.5 min for D-AP5 and 17.8 min for MK-801. The activity suppression was reversed following >40 min drug wash-out (n = 2). Electrical coupling through gap-junctions underlies generation of HFOs in the absence of chemical synaptic transmission [14], and the role of gap-junctions in HFOs has been pointed out using several experimental models [37], but some studies have questioned this relationship [11,30]. In this study, the possible involvement of gap-junctions on HFOs was examined by perfusing the slices with the gapjunction blocker carbenoxolone. The drug, at the concentrations of 0.1 mM produced a significant decrease in the rate of spontaneous events, from 17.1 ± 4.3 to 6.3 ± 1.4 events/min (n = 11, P < 0.05); however, it did not consistently affect the frequency (188.6 ± 16.2 Hz versus 185.6 ± 10 Hz, in control and drug conditions, respectively. Fig. 3C and E). At the concentrations of 0.2 mM carbenoxolone eventually abolished spontaneous activity (the rate was decreased from 17.85 ± 4.91 to 1.13 ± 0.46 events/min, n = 10, P < 0.0001, Fig. 2D and E). Carbenoxolone did not affect evoked NMDAR-mediated synaptic potentials (three slices, Fig. 3F). This study demonstrates that HFOs in the hippocampus can be sustained by synaptic transmission mediated solely by
C. Papatheodoropoulos / Neuroscience Letters 414 (2007) 197–202
199
Fig. 1. HFOs in the CA3 and CA1 regions of dorsal and ventral hippocampal slices. (A) Examples of field recordings from the CA3 st. pyramidale of a dorsal and a ventral slice (upper and lower panel, respectively). Continuous recordings of spontaneous activity (traces on the left) and single events (traces on the right) are shown. Note the lower rate of spontaneous events in the dorsal slice. Power spectra (inserts) illustrate the high-frequency peaks (200–250 Hz) revealed after fast fourier transformation. (B) Single events (traces on the left) recorded from DG and hilus of two different ventral slices. The corresponding power spectra are shown on the right. (C) Diagram where the rate of spontaneous events is plotted against the position of the slices along the dorsal–ventral axis of hippocampus. Note that the rate increases in ventral slices taken from progressively more extreme positions (positions 14–18, r = −0.73, P < 0.001) whereas the rate is similar among dorsal slices (positions 3–8). (D) Power spectra of CA3 (upper panel) and CA1 (lower panel) st. pyramidale recordings illustrating the higher frequency of activity in the CA3 compared to CA1 region. Inserts show wide-band (upper traces) and band pass (lower traces) recordings of single events. (E) Histograms of the high-frequency peaks observed in recordings from the CA3 and CA1 region. Note that most of the CA3 values were distributed at the range of 180–300 Hz, whereas CA1 presented an additional distribution at the range of 100–150 Hz. (F) Histograms of the CA3 high-frequency peaks in dorsal and ventral slices. (G) Example of CA1 evoked population spikes recorded from a slice in standard medium at the beginning of the experiment, i.e. before application of the drug cocktail (a) and at the end of the experiment, i.e. after three hours of washing out all drugs and reintroducing magnesium in the medium (b). Averages of four sweeps are shown. Stimulation artifacts are truncated.
NMDARs and that CA3 and CA1 display distinct features of such activity. Similarly to previous studies [5,11,26] HFOs were observed in the absence of fast GABAergic transmission. HFOs were observed in the CA3 and CA1 subregions in contrast to a recent study, in disinhibited slices, showing HFOs in the DG but not in CA3 and CA1, where paroxysmal events were seen [11].
This may be accounted for by the fact that excitation resting on non-NMDARs under complete blockade of inhibition [11] in combination with the large amount of recurrent excitatory collaterals in CA3 region [1] leads to exaggerated depolarizations which presumably disturb the generation of HFOs. In accord with this idea, HFOs have been observed in the disinhibited
200
C. Papatheodoropoulos / Neuroscience Letters 414 (2007) 197–202
Fig. 2. Both, CA3 and CA1 were sites of initiation of NMDA-dependent HFOs. (A) Simultaneous recordings (traces on the left column) of a ventral slice and corresponding power spectra (on the right), exemplifying the most frequent situation where oscillation was initiated in CA3 and propagated toward CA1 region. Dotted line is positioned at the start of CA3 activity. (B) An example of simultaneous recordings from CA3 and CA1 from a dorsal slice showing very small activity delay (∼2.5 ms) between the two regions. In this example, both regions displayed similar oscillation frequency. Note the large cross correlation of the two recordings (diagram on the right). (C) Representative example of the experimental cases where oscillation was initiated in the CA1 region. Recordings were obtained from a ventral slice. (D) Another example with activity initiation in CA1 region (a). On the same slice, activity was occasionally recorded in CA1 with only a very small signal in CA3 region. Data acquired from a dorsal slice. (E) An instance from simultaneous CA3 and CA1 recordings from another dorsal slice showing complete absence of activity in CA3 despite strong HFO in CA1 region. All recordings in the figure are band-pass filtered (80–550 Hz). However, all fast fourier transforms were obtained from unfiltered raw data.
CA1 minislice [26], in which the massive input from CA3 to CA1 was eliminated. Action potential bursts mediated by recurrent connections and intrinsic firing mechanisms are importantly involved in excitation-dependent HFOs [15,21]. Therefore, the recurrent connectivity and intrinsic firing properties, which are both different in CA3 and CA1 neurons [13], might explain to some extent the distinct oscillation frequencies seen in CA3 and CA1. The CA3 to DG propagation of activity, also previously shown [11,35], presumably resulted from the functionally uncovered, under disinhibition, CA3 to DG polysynaptic connections [35,36] with the crucial role of hilar neurons [7,35]. The independence of CA1 in generating HFOs, also previously observed [10,23,26,28], could be explained by the relatively higher content of NMDARs in CA1 [24] in combination with their involvement in CA1 polysynaptic activation under similar
to the present study conditions [9]. NMDARs in recurrent collaterals may provide a key element in recall memory mechanisms served by network oscillations [20]. Furthermore, NMDARs are strictly involved in memory processes carried out by specific hippocampal subregions [25,31]. Thus, the distinct frequency preference in the NMDAR-dependent HFOs displayed by the CA3 and CA1 hippocampal subregions in combination with the distinct NMDAR-dependent roles in memory processes taken by these subregions [22] corroborate the concept that information flow through hippocampus is subject to distinct processing by the discrete circuits of its subregions [21,32,38]. In line with previous reports [14,17,23,28], it is shown here that gap-junction blockade suppresses the generation of HFOs. Other studies however, have previously shown that gap-junction uncouplers failed to suppress HFOs induced either by disinhibition [11] or by tetanic stimulation [30]. It seems therefore
C. Papatheodoropoulos / Neuroscience Letters 414 (2007) 197–202
201
Fig. 3. Pharmacological examination of HFOs. (A and B) The antagonists of NMDARs D-AP5 and MK-801 consistently abolished spontaneous activity. Continuous recordings of spontaneous activity from CA3 (A) and CA1 (B) and the corresponding power spectra before (a) and during (b) drug application (black bars) are shown. Note that both drugs completely abolished spontaneous activity and the associated HFO in both CA3 and CA1 regions. (C and F) Effects of carbenoxolone on HFOs. (C and D) Continuous recordings of CA3 spontaneous activity (traces on the left) and the corresponding power spectra (diagrams on the right) from two slices. (C) The drug at the concentrations of 0.1 mM decreased the incidence of spontaneous potentials without significant effect on HFO frequency. At 0.2 mM (D) carbenoxolone strongly suppressed spontaneous potentials and accompanying HFOs (E (a), lower traces) but did not affect the evoked NMDAR-mediated synaptic response (E (b)) recorded from the same slice. F. Cumulative results showing the concentration-dependent effect of carbenoxolone on the rate of spontaneous activity.
that although electrical coupling is crucial in several models of high-frequency activity, in other models alternative mechanisms may contribute in synchronizing neuronal populations into coherent activity. It is also possible that the effects of carbenoxolone are mediated by a mechanism not involving gap junctions [33], although this seems unlikely given that the drug did not affect NMDAR-mediated responses. The higher incidence of spontaneous activity in ventral compared with dorsal slices is in line with previous data showing a positive correlation between neuronal excitability and the longitudinal axis of the hippocampus in the dorso-ventral direction [3,12,16,29]. The ventral compared to dorsal hippocampus
exhibits a higher probability of transmitter release at its excitatory synapses [27] and higher levels of extracellular glutamate [19]. These may contribute to the higher excitability of the ventral hippocampus observed in the present study, expressed as a higher incidence of spontaneous events, taking into account that extracellular glutamate enhances neuronal excitability by tonically activating NMDARs [34]. In conclusion, the present findings reveal NMDARs as an alternative source of excitatory drive in the generation of network HFOs. It seems that hippocampal circuits can achieve a high-frequency oscillatory state by engaging several distinct mechanisms. Furthermore, the oscillations in CA3 and CA1
202
C. Papatheodoropoulos / Neuroscience Letters 414 (2007) 197–202
hippocampal subregions expressed distinct characteristics suggesting that the microcircuitry of each region produces and sustains activity of this type in a distinct way. Acknowledgments Study funded by the European Social Fund (ESF), Operational Program for Educational and Vocational Training II (EPEAEK II), and particularly the Program PYTHAGORAS II. I wish to thank Dr. George Kostopoulos for helpful comments on the manuscript. References [1] D.G. Amaral, N. Ishizuka, B. Claiborne, Neurons, numbers and the hippocampal network, Prog. Brain Res. 83 (1990) 1–11. [2] C. Bonansco, W. Buno, Cellular mechanisms underlying the rhythmic bursts induced by NMDA microiontophoresis at the apical dendrites of CA1 pyramidal neurons, Hippocampus 13 (2003) 150–163. [3] C. Borck, J.G. Jefferys, Seizure-like events in disinhibited ventral slices of adult rat hippocampus, J. Neurophysiol. 82 (1999) 2130–2142. [4] A. Bragin, J. Engel Jr., C.L. Wilson, I. Fried, G. Buzsaki, High-frequency oscillations in human brain, Hippocampus 9 (1999) 137–142. [5] A. Bragin, I. Mody, C.L. Wilson, J. Engel Jr., Local generation of fast ripples in epileptic brain, J. Neurosci. 22 (2002) 2012–2021. [6] G. Buzsaki, Z. Horvath, R. Urioste, J. Hetke, K. Wise, High-frequency network oscillation in the hippocampus, Science 256 (1992) 1025–1027. [7] G.C. Carlson, Role of hilar neurons in the propagation of fast-ripples from CA3 into the dentate gyrus: a fast voltage-sensitive dye imaging study, Soc. Neurosci. Abstr. (2004) (No. 853.2.). [8] L.L. Colgin, Y. Jia, J.M. Sabatier, G. Lynch, Blockade of NMDA receptors enhances spontaneous sharp waves in rat hippocampal slices, Neurosci. Lett. 385 (2005) 46–51. [9] V. Crepel, R. Khazipov, Y. Ben-Ari, Blocking GABA(A) inhibition reveals AMPA- and NMDA-receptor-mediated polysynaptic responses in the CA1 region of the rat hippocampus, J. Neurophysiol. 77 (1997) 2071–2082. [10] J. Csicsvari, H. Hirase, A. Czurko, A. Mamiya, G. Buzsaki, Fast network oscillations in the hippocampal CA1 region of the behaving rat, J. Neurosci. 19 (1999) RC20. [11] M. D’Antuono, P. de Guzman, T. Kano, M. Avoli, Ripple activity in the dentate gyrus of dishinibited hippocampus-entorhinal cortex slices, J. Neurosci. Res. 80 (2005) 92–103. [12] M. Derchansky, E. Shahar, R.A. Wennberg, M. Samoilova, S.S. Jahromi, P.A. Abdelmalik, L. Zhang, P.L. Carlen, Model of frequent, recurrent, and spontaneous seizures in the intact mouse hippocampus, Hippocampus 14 (2004) 935–947. [13] J. Deuchars, A.M. Thomson, CA1 pyramid-pyramid connections in rat hippocampus in vitro: dual intracellular recordings with biocytin filling, Neuroscience 74 (1996) 1009–1018. [14] A. Draguhn, R.D. Traub, D. Schmitz, J.G. Jefferys, Electrical coupling underlies high-frequency oscillations in the hippocampus in vitro, Nature 394 (1998) 189–192. [15] V.I. Dzhala, K.J. Staley, Mechanisms of fast ripples in the hippocampus, J. Neurosci. 24 (2004) 8896–8906. [16] M. Gilbert, R.J. Racine, G.K. Smith, Epileptiform burst responses in ventral vs dorsal hippocampal slices, Brain Res. 361 (1985) 389–391. [17] S.J. Gladwell, J.G. Jefferys, Second messenger modulation of electrotonic coupling between region CA3 pyramidal cell axons in the rat hippocampus, Neurosci. Lett. 300 (2001) 1–4. [18] M. Hausser, N. Spruston, G.J. Stuart, Diversity and dynamics of dendritic signaling, Science 290 (2000) 739–744.
[19] H. Hortnagl, M.L. Berger, G. Sperk, C. Pifl, Regional heterogeneity in the distribution of neurotransmitter markers in the rat hippocampus, Neuroscience 45 (1991) 261–272. [20] O. Jensen, J.E. Lisman, Theta/gamma networks with slow NMDA channels learn sequences and encode episodic memory: role of NMDA channels in recall, Learn. Mem. 3 (1996) 264–278. [21] R.P. Kesner, I. Lee, P. Gilbert, A behavioral assessment of hippocampal function based on a subregional analysis, Rev. Neurosci. 15 (2004) 333–351. [22] I. Lee, R.P. Kesner, Differential contribution of NMDA receptors in hippocampal subregions to spatial working memory, Nat. Neurosci. 5 (2002) 162–168. [23] N. Maier, V. Nimmrich, A. Draguhn, Cellular and network mechanisms underlying spontaneous sharp wave-ripple complexes in mouse hippocampal slices, J. Physiol. 550 (2003) 873–887. [24] D.T. Monaghan, D. Yao, H.J. Olverman, J.C. Watkins, C.W. Cotman, Autoradiography of d-2-[3H]amino-5-phosphonopentanoate binding sites in rat brain, Neurosci. Lett. 52 (1984) 253–258. [25] K. Nakazawa, M.C. Quirk, R.A. Chitwood, M. Watanabe, M.F. Yeckel, L.D. Sun, A. Kato, C.A. Carr, D. Johnston, M.A. Wilson, S. Tonegawa, Requirement for hippocampal CA3 NMDA receptors in associative memory recall, Science 297 (2002) 211–218. [26] V. Nimmrich, N. Maier, D. Schmitz, A. Draguhn, Induced sharp wave-ripple complexes in the absence of synaptic inhibition in mouse hippocampal slices, J. Physiol. 563 (2005) 663–670. [27] C. Papatheodoropoulos, G. Kostopoulos, Dorsal–ventral differentiation of short-term synaptic plasticity in rat CA1 hippocampal region, Neurosci. Lett. 286 (2000) 57–60. [28] C. Papatheodoropoulos, G. Kostopoulos, Mechanisms of generation of a novel rhythmical network activity in ventral hippocampal slices, Soc. Neurosci. Abstr. (2003) (No. 582.8.). [29] C. Papatheodoropoulos, C. Moschovos, G. Kostopoulos, Greater contribution of N-methyl-d-aspartic acid receptors in ventral compared to dorsal hippocampal slices in the expression and long-term maintenance of epileptiform activity, Neuroscience 135 (2005) 765–779. [30] B. Poschel, U. Heinemann, A. Draguhn, High frequency oscillations in the dentate gyrus of rat hippocampal slices induced by tetanic stimulation, Brain Res. 959 (2003) 320–327. [31] T. Rajji, D. Chapman, H. Eichenbaum, R. Greene, The role of CA3 hippocampal NMDA receptors in paired associate learning, J. Neurosci. 26 (2006) 908–915. [32] R.G. Robertson, E.T. Rolls, P. Georges-Franois, Spatial view cells in the primate hippocampus: effects of removal of view details, J. Neurophysiol. 79 (1998) 1145–1156. [33] N. Rouach, M. Segal, A. Koulakoff, C. Giaume, E. Avignone, Carbenoxolone blockade of neuronal network activity in culture is not mediated by an action on gap junctions, J. Physiol. 553 (2003) 729–745. [34] P. Sah, S. Hestrin, R.A. Nicoll, Tonic activation of NMDA receptors by ambient glutamate enhances excitability of neurons, Science 246 (1989) 815–818. [35] H.E. Scharfman, Evidence from simultaneous intracellular recordings in rat hippocampal slices that area CA3 pyramidal cells innervate dentate hilar mossy cells, J. Neurophysiol. 72 (1994) 2167–2180. [36] H.E. Scharfman, Conditions required for polysynaptic excitation of dentate granule cells by area CA3 pyramidal cells in rat hippocampal slices, Neuroscience 72 (1996) 655–658. [37] R.D. Traub, Fast oscillations and epilepsy, Epilepsy Curr. 3 (2003) 77–79. [38] A. Treves, Computational constraints between retrieving the past and predicting the future, and the CA3–CA1 differentiation, Hippocampus 14 (2004) 539–556. [39] H. Yoshimura, N. Kato, M. Honjo, T. Sugai, N. Segami, N. Onoda, Agedependent emergence of a parieto-insular corticocortical signal flow in developing rats, Brain Res. Dev. Brain Res. 149 (2004) 45–51.