The adult rat hippocampal slice revisited with multi-electrode arrays

BR A IN RE S EA RCH 1 0 96 ( 20 0 6 ) 7 0 –84

a v a i l a b l e a t w w w. s c i e n c e d i r e c t . c o m

w w w. e l s e v i e r. c o m / l o c a t e / b r a i n r e s

Research Report

The adult rat hippocampal slice revisited with multi-electrode arrays Esther-Marie Steidl a,c , Estelle Neveu a , Daniel Bertrand b , Bruno Buisson a,c,⁎ a

TROPHOS SA, case 931, Parc Luminy Biotech Entreprises F-13288 Marseille cedex 09, France Department of Neuroscience, Faculty of Medicine, 1 rue Michel Servet, CH-1211 Geneva 4, Switzerland c Neuroservice, INMED, Parc Scientifique de Luminy, 13273 Marseille cedex 09, France b

A R T I C LE I N FO

AB S T R A C T

Article history:

The multi-electrode arrays (MEA) technology for the recording of brain slices is available for

Accepted 10 April 2006

more than 10 years. However, despite its relative simplicity, this recording technique is not

Available online 23 May 2006

widely used in academic or pharmaceutical research laboratories. We illustrate here that MEA provide multiple possibilities to investigate some network physiological properties as well as

Keywords:

to evaluate the pharmacological effects of compounds. We first document that MEA allow to

Multi-electrode array

trigger and to record conventional FP which are inhibited by the block of action potential

Hippocampal slice

propagation (with 500 nM TTX). FP recorded with MEA are sensitive to ionic substitutions, to

Synaptic transmission

ionotropic glutamate receptor antagonists (CNQX or NBQX) and to energetic failure. Second,

Pharmacology

we illustrate that different “classical” protocols (paired-pulse, LTP, chemical LTD), revealing

GABA-A receptor

synaptic plasticity mechanisms, could be performed. Third, we document that MEA allow

Steroid

spatial and temporal discriminations for the effects of known pharmacological compounds such as competitive antagonist (gabazine, bicuculline) and allosteric modulators (steroids) of GABAA receptors. In conclusion, we illustrate that MEA recordings of adult rat hippocampal slices constitute a powerful and sensitive system to evaluate the effect of molecules on basic synaptic propagation/transmission and on synaptic plasticity processes. © 2006 Elsevier B.V. All rights reserved.

1.

Introduction

More than 200 years ago, the fundamental experiments of Galvani revealed that “bioelectric forces” exist in living tissues such as nerve and muscles. In 1843, Du Bois-Reymond demonstrated that the “nerve impulse” is a wave-like propagation of negative charges in the nerve trunk. But it took almost one additional century before Huxley proposed the first model of the nerve axon (1938). The demonstration of the ionic mechanisms supporting nerve conduction was achieved by Hodgkin and Huxley (1952a,b) in 1952. Since these pioneering studies, the glass electrode has been

considered as the “standard” for electrophysiological recordings, and the refinement of this technique has culminated by the development of the patch-clamp (Hamill et al., 1981). However, alternative approaches for the non-invasive recording of neurons have been developed: multi-electrode arrays (MEA) made of metal electrodes or semi-conductors (Hutzler and Fromherz, 2004) and two-photon imaging technologies. One of the main problems when using metal electrodes to record extracellular signals is to minimize redox phenomena that could interfere with biological processes and signals. Electrodes of MEA are made of inert metal such as indium tin oxide (ITO), platinum or gold that could be additionally coated

⁎ Corresponding author. Present address: Neuroservice, INMED, Parc Scientifique de Luminy, 13273 Marseille cedex 09, France. Fax: +33 4 91 82 82 85. E-mail address: [email protected] (B. Buisson). 0006-8993/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.brainres.2006.04.034

BR A IN RE S E A RCH 1 0 96 ( 20 0 6 ) 7 0 –8 4

71

Fig. 1 – Amplitude of Field Potentials and Fiber Volley as a function of stimulus intensity (either current or voltage). Slices were stimulated either with symmetric bipolar current (A, B) or voltage (C, D) pulses in the CA3 region. Field Potentials (FP) or Fiber Volley (FV) was recorded in CA1 (stratum radiatum). Mean values of FP or FV amplitudes (in μV) were plotted as a function of the stimulation intensity. In the same slices (A or B), increasing current stimuli were first applied between one MEA electrode and the reference electrode (squares) and then between two adjacent MEA electrodes (triangles) including the electrode that was used initially against the ground. Evoked FP and FV were twice bigger when the current stimulus is applied between two MEA electrodes. A good linearity of the responses versus stimuli is observed for current intensity between 50 and 600 μA (A, B). For stimuli >600 μA, a clear rectification appears for the FV amplitude which might indicate that whole fibers have been recruited. In the same slices (C or D), increasing voltage stimuli were first applied between one MEA electrode and the reference electrode (squares) and then between two adjacent MEA electrodes (triangles) including the electrode that was used initially against the ground. Voltage stimuli (C) did not evoke responses as large as the ones observed with current stimuli (A), but the voltage–response relationship is linear over the range of values investigated (C). As previously observed for current stimulations, the amplitude of the FV is larger for bipolar stimulations (D). (A and C) Mean of 6 electrodes/slice and 2 slices for each protocol. (B and D) Mean of 8 electrodes in 3 independent slices. Error bars correspond to SEM.

by a biologically inert metal layer of black platinum, or titanium nitride. There are now a growing number of publications that demonstrate that neuron cultures as well as nervous tissue slices could be recorded and stimulated with MEA. If technical and methodological works describing the realization and the basic use of MEA are now well documented, their use in extensive biological studies has been documented in only a few studies (Meister et al., 1991; Tscherter et al., 2001; Shimono et al., 2002; Wirth and Luscher, 2004). We present here multiple ways for the use of MEA in the rat adult hippocampal slice to demonstrate that such technique could be easily used for physiological and pharmacological studies. Finally, and may be of most importance, we would like to shed light on the fact that MEA allow multi-sites

recording within a single slice enabling powerful statistical analysis and observation of region-specific phenomenon.

2.

Results

2.1.

Current versus voltage stimulation with MEA

MEA electrodes display the unique advantage to be used either as active or as passive electrodes. To document the signals that could be triggered by the MEA, we have performed in the same slices different stimulating protocols based on current or voltage pulses applied between one electrode and the ground reference electrode (monopolar) or between two adjacent

72

BR A IN RE S EA RCH 1 0 96 ( 20 0 6 ) 7 0 –84

electrodes (bipolar), one of the two electrode being previously used for stimulating against the ground. We have chosen to record Field Potentials (FP) and Fiber Volley (FV) in the CA1 region while stimulating the Schaeffer collaterals in the CA3/ CA1 region; a very classical protocol for the hippocampal slice. When increasing bipolar current pulses (see Experimental procedures) are applied, and independently from the configuration used, we observe that the threshold to evoke detectable FP is around 100 μA (Fig. 1A). However, and for equivalent current intensities, FP are about twice as bigger when the current stimulus is passed between two adjacent electrodes (Fig. 1A). For a ± 600 μA pulse, the amplitude mean values of the FP are of −560 ± 78 μV when stimulating between one electrode and the ground and of −1022 ± 133 μV when stimulating between two adjacent electrodes. Stimuli– response curves reveal an overall linear relationship for intensities ranging between 100 and 600 μA; for intensities >600 μA, the stimulus–response curve displays a “rectification” since maximal responses amplitude is approached. Because of technical limits of the stimulus generator, higher intensities could not be applied. Differences in FP amplitudes according to the stimulation protocol (monopolar versus bipolar) could result from differential efficacy in recruiting afferent fibers. This hypothesis is exemplified by measurements of FV amplitude. As presented on Fig. 1B, the FV amplitude is twice as bigger with bipolar stimulations when compared to equivalent monopolar stimulations. This phenomenon indicates that bipolar stimuli recruit more afferent fibers than monopolar ones. Similar protocols performed with bipolar voltage pulses reveal that the threshold to evoke detectable responses was around 500 mV and that evoked FP display a linear relationship over the range of stimulating potentials investigated (500–4000 mV; Fig. 1C). There are fewer differences in the amplitudes recorded when using monopolar or bipolar stimulations (Fig. 1C). At the maximal voltage intensity (4000 mV), the mean amplitude responses are of −564 ± 92 μV when stimulating between one electrode and the ground and of −789 ± 135 μV when stimulating between two adjacent electrodes. Because of the technical limits of the stimulus generator, voltage amplitude over 4000 mV could not be investigated. As observed for current stimulations, FV is larger when slices are stimulated between two adjacent electrodes. We conclude that FP can be evoked by using either current (100–600 μA) or voltage (500–4000 mV) pulses and that the response amplitudes are very closed independently of the nature of the stimulus and of configurations (monopolar or bipolar). However, larger responses could be triggered with bipolar stimuli in the current mode. For the next experiments, and if not specified, we have used a monopolar current stimulation of ±300 μA that evokes half-maximal intensity responses.

2.2. Field potentials are sensitive to ionic substitutions, ion channel antagonists and energy failure To demonstrate that FP recorded by the MEA electrodes corresponds to biological signals, we have realized series of experiments using reference pharmacological compounds and ionic substitutions. Perfusion of slices with the selective “fast”

sodium channels antagonist Tetrodotoxin (TTX), 500 nM results in the fast decrease of the FP amplitude (74.7 ± 3.3%; Figs. 2A and D), an effect that could be attributed to the inhibition of spikes propagation along the axons of Schaeffer collaterals. When the ACSF sodium concentration is decreased by 50% (substitution of sodium ions by saccharose) the FP amplitude is reduced by 47.5 ± 0.5% (Fig. 2D). Half inhibition of the amplitude shall be attributed to a 50% decrease of the charges flowing through Na+-permeable AMPA (α-amino-3hydroxy-5-methyl-4-isolaxone propionate)/kainate receptors and a subsequent 50% decrease of membrane depolarization. When the slice is bathed with a “low calcium–high magnesium” solution, a classical protocol to decrease neurotransmitter release (Doller and Weight, 1982; Chang and Greenough, 1984), once again the FP are strongly inhibited by 68.7 ± 1.1% (Fig. 2B). Because 3 μM NBQX (6-nitro-7-sulphamobenzoquinoxaline-2,3-dione) and 50 μM CNQX (6-cyano-7-nitroquixaline-2,3-dione) decrease FP amplitude respectively from 74.9 ± 4.2% and 73.8 ± 4.5% (Figs. 2C and D) most of the field potential signal might result from depolarization triggered by AMPA/kainate receptors and shall be considered as excitatory post-synaptic FP (fEPSPs). It is of value to observe that in the contrary to TTX, which inhibit action potential propagation along Schaeffer collaterals, NBQX selectively inhibits synaptic transmission. Thus, the FV is still observed following NBQX inhibition (data not shown). Finally we have applied on slices carbonyl cyanide-p-trifluoromethoxyphenyl-hydrazon (FCCP), a cyanide derivative that provokes a complete collapse of the mitochondrial proton gradient and a rapid energetic failure of the neurons. The field potential amplitude is reduced by 81.8 ± 2.5% after 30-min perfusion of 10 μM FCCP (Figs. 2B and D). This effect might result from important energetic failure following mitochondria depolarization. We conclude from these experiments that MEA recorded FP depend on action potential propagation, synaptic transmission and sodium influx through AMPA/Kainate receptors.

2.3. Presynaptic facilitation revealed by paired-pulse protocols One form of short-term synaptic plasticity involves an increase of neurotransmitter release at presynaptic boutons, a mechanism that depends mainly on intracellular calcium rises following the action potential spreading. The paradigm is the more calcium, the more neurotransmitter release. Then, when two identical pulses are applied with a 50-ms interval (paired-pulses), once the second action potential arrives at the synaptic bouton, the calcium level has not returned back to its baseline value and it rises to a higher concentration. As a consequence, more neurotransmitter is released and the field potential amplitude is increased. Data presented in Fig. 3 demonstrate that substantial and steady potentiation of the second field potential can be observed by using a paired-pulses protocol. Mean potentiation (ratio between peak no. 2 and peak no. 1) is of 45 ± 3% (Fig. 3B) but can reach up to 200–300%. The potentiation observed between the two peaks is stable over 1-h time scale (data not shown). It is of value to note that such potentiation is observed only for low to moderate intensity values of stimulation (from 150 up to 300 μA).

BR A IN RE S E A RCH 1 0 96 ( 20 0 6 ) 7 0 –8 4

73

Fig. 2 – Pharmacological and physiological validation of field potential signals. Panels A–C present representative traces of FP recorded in CA1 before and 10 min after drug perfusion. Almost complete inhibition of the FP was observed after inhibiting spikes propagation with 500 nM TTX (A) or following energetic failure (by poisoning oxidative ATP production in mitochondria with the protonophore FCCP, (B)) or by blocking glutamatergic transmission with the selective AMPA/ kainate receptors antagonist NBQX (C). A stable 10-min baseline was recorded each time before drug perfusion. (D) Mean inhibition values of FP observed with different pharmacological agents or by ionic substitutions (see the text for details); 500 nM TTX (n = 16, 3 slices), half sodium decrease (n = 9, 3 slices), low calcium–high magnesium (n = 9, 3 slices), 50 μM CNQX (n = 16, 4 slices), 3 μM NBQX (n = 12, 2 slices), 10 μM FCCP (n = 23, 3 slices). Error bars correspond to SEM.

Fig. 3 – Facilitation of synaptic transmission is observed with paired-pulse protocols. Two submaximal current pulses (±150 μA) are applied at 50-ms interval in the CA1 region of the hippocampus every 30 s. As expected, the second pulse triggers a field potential of larger amplitude on each of the recorded electrodes. A representative recording is presented in panel (A). The second response is potentiated by almost 50%, and the signals stability is robust over minutes (B, n = 39, 6 slices). Error bars are included in symbols and represent SEM.

74

BR A IN RE S EA RCH 1 0 96 ( 20 0 6 ) 7 0 –84

2.4. Long-term potentiation is induced at CA3/CA1 synapses and is blocked by D-AP5 The possibility to investigate in vitro synaptic plasticity mechanisms represent one of the most attractive activity of neurobiologists that aim to understand the molecular and cellular mechanisms of different form of information processing and storage. Out of the many preparations and protocols investigated so far, long-term potentiation (LTP) constitutes the most widely used paradigm to investigate some forms of synaptic plasticity. We have performed very classical 100-Hz stimulations at CA3/CA1 synapses and, as illustrated in Fig. 4A, a robust and persistent LTP can be induced in the stratum radiatum region. The mean increase of FP a few minutes after high-frequency stimulation (HFS) is around 60% and reaches 80 ± 9% 2 h after HFS. Because LTP at CA3/CA1 synapses depends on NMDA receptors activation, it should be blocked by selective NMDA receptor antagonists such as D-AP5. Perfusion of 30 μM D-AP5 has no effect on basal synaptic transmission, but it considerably reduces LTP induction (see Figs. 4A–C). Thirty minutes after HFS protocol, FP are increased by 68 ± 9% in control conditions whereas no potentiation is observed if 30 μM DAP-5 are pre- and co-perfused (increase of 5 ± 2%). It is of value to note that LTP amplitude is variable according to the location of the electrode in the slice (see below) and the age of the rat. Range of potentiation could oscillate between 30 and 300%. Years of experiments with MEA lead us to the conclusion that mean LTP amplitude recorded by each electrode in one defined layer of a slice is very similar for all of the slices prepared from a single hippocampus. As a consequence, all experiments presented in this work (to compare LTP in different conditions) were realized the same(s) day(s), by alternating controls and drugexposed slices. Another important feature revealed by MEA recordings is that LTP in stratum pyramidale is of much larger amplitude than in stratum radiatum (Fig. 4D). Indeed, FP increase by 158 ± 16% in stratum pyramidale immediately after HFS, whereas only a 50 ± 5% potentiation is measured in stratum radiatum. Such a difference in LTP amplitudes according to the layer of the hippocampus persists over the time. Twenty minutes after LTP induction, the mean potentiation is of 139 ± 16% and of 58 ± 5% in stratum pyramidale and in the stratum radiatum, respectively.

2.5. Chemical induction of long-term depression with the mGluR1/mGluR5 agonist DHPG Another form of synaptic plasticity is long-term depression (LTD) that corresponds to a reduction in synaptic transmission efficiency. LTD could be induced in the CA1 area of the hippocampus either by low-frequency stimulations (Mockett et al., 2002) or by chemicals such as the selective mGluR1/ mGluR5 glutamate receptors agonist (RS)-3,5-dihydroxyphenylglycine (DHPG) (Palmer et al., 1997; Fitzjohn et al., 2001). As illustrated in Fig. 5, perfusion of 30 μM DHPG on the slice induced a 32.2 ± 0.5% steady-state inhibition of FP that can be maintained as long as the compound is perfused and that persists several minutes after washout (data not shown).

Fig. 4 – Inhibition of LTP by the NMDA receptor antagonist D-AP5 and region-specific features of LTP. Using a 100 Hz stimulation protocol a robust and persistent LTP could be induced at CA3/CA1 synapses. (A) Representative FP recorded in the stratum radiatum before and after LTP induction. FP amplitudes were normalized for each electrode to the mean control value (average of amplitude values over the 10 minute baseline period). Normalized FP values are plotted as a function of the time and superimposed for comparison between control conditions (squares, n = 5, 1 slice) and when 30 μM D-AP5 is pre-perfused ten minuets before LTP induction (triangle, n = 7, 1 slice). Individual recordings presented in (B) and (C) are indicated by the numbers between brackets above data point traces. Panel (B) displays FP recorded by a single electrode in stratum Radiatum 10 min before and 30 min after 100 Hz LTP induction. Panel (C) illustrates the full inhibition of LTP induction by 30 μM D-AP5 in another slice of the same recording session. Panel (D) illustrates the difference in LTP amplitude between different layers of the hippocampus: the stratum radiatum and stratum pyramidale. Potentiation in stratum pyramidale (SP; open circle, n = 7, 2 slices) is almost twice larger than the one observed in the stratum radiatum (SR; open circle, n = 9, 2 slices). Such a regional difference in LTP amplitude has been observed in whole of the slices investigated so far. Error bars correspond to SEM.

BR A IN RE S E A RCH 1 0 96 ( 20 0 6 ) 7 0 –8 4

Fig. 5 – Long-term depression induced by DHPG, a metabotropic glutamate agonist. LTD could be induced and maintained at CA3/CA1 synapses by the continuous perfusion of the selective mGluR1/mGluR5 agonist DHPG at 30 μM. (A) The FP amplitude was recorded over a 10-min period before perfusion of DHPG for 20 min. Data points corresponding to the mean of normalized individual amplitude are plotted as a function of time (n = 9, 1 slice). Error bars correspond to SEM. Single traces of FP are displayed in panel (B).

2.6. Gabazine displays concentration- and spatial-dependent effects in the CA1 region There is now a consensus about the idea that two types of GABAergic neurotransmissions exist in the adult hippocampus. “Phasic” transmission corresponds to the classical view of neurotransmission, that is, GABA is released and diffuses in the ms time scale to activate post-synaptic GABAA receptors embedded in the post-synaptic membranes. “Tonic” transmission is the result of the continuous opening of high-affinity extrasynaptic GABAA receptors where GABA can slowly diffuse from synaptic cleft. Both types of GABAA receptors shall be involved in the control of network excitability but on the basis of different mechanisms (Semyanov et al., 2004; Cossart et al., 2005; Farrant and Nusser, 2005). Gabazine (SR-95531) is a competitive antagonist of GABAA receptors that has the unique property to selectively inhibit the different types of GABAA receptors, depending on its concentration. At a submicromolar concentration, gabazine inhibits only “phasic” GABAA receptors, whereas for concentrations over 10 μM, it can blocks both “phasic” and “tonic” GABAA receptors (Bai et al., 2001; Stell and Mody, 2002; Yeung

75

et al., 2003). When the adult hippocampal slice is exposed to 500 nM gabazine FP recorded within the stratum radiatum are not affected (Fig. 6). In contrast, FP increase in the stratum radiatum close to the stratum pyramidale layer (until 56% for electrode 57) and a weak epileptiform activity can be observed (Fig. 6). This small potentiation might result from the disinhibition of pyramidal cells where the phasic inhibition by GABAergic interneurons could be blocked by a low concentration of gabazine (Bai et al., 2001). In addition, beginning of oscillations was only observed on the electrodes close to the pyramidale cell layer. Our data suggest that selective inhibition of “phasic” GABAA-mediated currents has only a small impact on the neurons close to the stratum pyramidale layer. When gabazine is perfused at 50 μM, no major modification of the FP amplitude is observed within the stratum radiatum but a sustained epileptiform activity appears (Fig. 7). In the region closed to the stratum pyramidale layer, large increase of FP amplitude is observed together with epileptiform activity (615% increase is observed on electrode 42; see Fig. 7D). These effects are very similar to the ones that can be observed with bicuculline, another GABAA receptor antagonist that inhibits both types of GABAA receptors. Consistent with a block of inhibitory inputs, bicuculline highly increases FP in the more external area of stratum radiatum and induces strong epileptiform activity. There is a 512% increase of the FP amplitude at electrode 33 (see Figs. 8A–D). However, and as observed with 50 μM gabazine, bicuculline has almost no effect on the FP recorded in the more internal area of this layer as well as those recorded in stratum lacunosum-moleculare. As an illustration, FP recorded by electrode 64 in Fig. 8 are increased by 12%. Only a slight epileptiform activity can be observed on these electrodes.

2.7. Slight effects of neurosteroids in the CA1 region are detected with the MEA Differential effects of gabazine are observed in the hippocampal slice. They might result from the selective block of different type of receptors: “tonic” versus “phasic” GABAA receptors. Since “tonic” GABAA receptors are reported to be very sensitive to the effect of neurosteroids (Stell et al., 2003; Bianchi and MacDonald, 2004; Olsen et al., 2004), we have decided to investigate if the effects of such molecules could be detected using MEA recordings. Tetrahydro-progesterone (3α, 5α-TH-PROG or allopregnanolone) and tetrahydro-deoxycorticosterone (3α, 5α-THDOC) are endogenous brain steroids which are known as the most potent allosteric modulators of GABA A receptors (Majewska et al., 1986; Belelli and Lambert, 2005). Perfusion of saturating concentrations of 3α, 5α-TH-PROG (100 nM) or 3α, 5αTH-DOC (100 nM) induces a slight, but significant increase of the mean FP amplitude recorded in the stratum Radiatum (Fig. 9A). At 100 nM, 3α, 5α-TH-PROG and 3α, 5α-TH-DOC increase the mean FP amplitude by 12.1 ± 0.9% and 8.77 ± 0.83%, respectively (Fig. 9A). No further potentiation could be observed for 1 μM concentrations (data not shown). The effect of 3α, 5α-TH-PROG is further enhanced and revealed when performing 100 Hz LTP protocols. In such conditions, a 100% increase of the “late” phase of LTP is observed (data not shown). The role of GABAA receptors in such effects is confirmed by the activity of two

76

BR A IN RE S EA RCH 1 0 96 ( 20 0 6 ) 7 0 –84

Fig. 6 – Selective inhibition of the GABAA-mediated phasic transmission by low concentrations of SR-95531 (gabazine) increases field potential amplitude only in the pyramidal layer. (A) Localization of the stimulating and recording electrodes in the layers of the hippocampal slice. The stimulating electrode is indicated in white. The recording electrodes are surrounded by a white line and traces presented in the other panels respect their geographical position within the slice. Time course of FP recorded before (B) and 20 min after the perfusion of 500 nM gabazine (C). At this submicromolar concentration, gabazine is known for selectively inhibiting GABAA-mediated phasic transmission. Individual electrodes FP plotted as a function of time are presented in panel (D). Note that only field potential in the top row, located in the pyramidal layer, is affected by gabazine (statistically significant, P < 0.001). Arrows indicate the time when gabazine perfusion started into the MEA chamber. These data are representative of three slices. synthetic steroids, alphaxalone and ganaxalone, the latest being a non-metabolized analog of 3α, 5α-TH-PROG. Saturating concentrations of alphaxalone (3 μM) and ganaxalone (1 μM) increase the mean FP amplitudes by 7.2 ± 0.9% and 15.6 ± 1.1%, respectively (Figs. 9A–D). Because potentiation of FP was more consistent with ganaxolone (may be attributable to its metabolic stability), the following experiments were performed with this synthetic steroid. The potentiation induced by steroids is small, and it could be objected that it could be an artefact linked either with the use of the MEA recording system or with signal processing (amplitude versus slope measurements). Most electrophysiological experiments performed with extracellular recordings rely on “slope factor” calculation and analysis. We have then verified that plot of the slope factor as a function of time allows revealing the steroid effect. Slope and amplitude parameters of FP were calculated in parallel for signals recorded at MEA electrodes. Results obtained with a representative slice are presented in Fig. 9. On average in this slice, the increase of the FP amplitude is of 20.8 ± 1.5%, whereas the slope factor is potentiated by 45.6 ± 6.8%. Then, the ganaxolone effect is, indeed, better revealed and confirmed by analysis of the slope factor parameter.

One can object that potentiations observed with steroids could be a consequence of “non-selective” stimulation of the neuronal network because monopolar stimulations were applied between one MEA electrode and the reference electrode. Then spread of the current between the MEA electrode and the reference electrode is much larger than the current that can pass between the two electrodes of a classical twisted wire. To address this issue, we have performed experiments based on bipolar stimulations (applied between two adjacent MEA electrodes, separated by 200 μm, which are closer than the ones of a twisted wire). In such conditions, we are still able to observe the potentiation of FP induced by ganaxolone (data not shown). Finally, we have verified the profile of signals that are recorded simultaneously with a MEA electrode and a glass electrode placed at the surface of the slice, just above the selected MEA recording electrode. The silver wire of the glass electrode was directly connected to one of the non-used amplifiers of the MEA head-stage and the amplifier disconnected from its MEA electrode. As illustrated in Fig. 10, no difference in the signals time course could be observed between the two types of electrodes and the ganaxolone effect is observed simultaneously on both sides of the slice

BR A IN RE S E A RCH 1 0 96 ( 20 0 6 ) 7 0 –8 4

77

Fig. 7 – Inhibition of phasic and tonic GABAA receptors by 50 μM gabazine lead to a strong increase of field potentials recorded in the stratum radiatum. Distribution of recording electrodes beneath the different layers of hippocampal slices is shown in the photograph (A). The stimulating electrode is white shaded and recorded electrodes are surrounded by a white line. Shape of FP before (B) and after 20-min perfusion of 10 μM gabazine on the slice (C) are illustrated for each electrode and plots of FP amplitudes as a function of time are presented in D (arrows indicate beginning of gabazine perfusion).

with the two types of electrodes (7.4 and 7.5% increase of FP at the MEA and glass electrode, respectively). In conclusion, we can say that a small but significant effect of steroids is measured in the CA1 area with the MEA recording system. This potentiation does not correspond to an artefact but shall be attributed to a modulation of network excitability following positive allosteric modulation of some GABAA receptors.

3.

Discussion

From a very practical point of view MEA recordings offer many advantages: – No need of micromanipulators to position the electrodes in the tissue; the slice is “sealed” on multiple electrodes that could be used either for stimulation or recording.

– MEA recordings are less sensitive to mechanical vibrations than glass electrode recordings. – MEA could be used several times when adequately cleaned after each recording session. – 3D-tip-shaped MEA allow to pass trough the dead cell layer and to record closer to living neurons within the slice. – Parallel recordings at multiple electrode sites in a single slice providing the opportunity. ○ to observe region-specific effects. ○ to increase individual data points and to improve statistical analysis. The present work illustrates that MEA are equivalent to glass electrodes to record acute brain slices. When looking at the amplitude of the responses that could be recorded with the MEA (from 0.2 up to 3 mV), we observe that signals are in the range of FP that could be recorded with standard

78

BR A IN RE S EA RCH 1 0 96 ( 20 0 6 ) 7 0 –84

Fig. 8 – Bicuculline induces the same effects than 50 μM gabazine. (A) Localization of the stimulating and recording electrodes in the layers of the hippocampal slice. The stimulating electrode is indicated in white. The recording electrodes are surrounded by a white line and traces presented in the other panels with respect to their geographical position within the slice. Time course of FP recorded before (B) and 20 min after the perfusion of 20 μM bicuculline (C), a selective GABAA receptors antagonist. Large amplitude oscillations as well as huge amplitude increase could be observed at some electrodes while other did not show major changes. Individual electrode FP plotted as a function of time are presented in (D). When oscillations occurred, amplitude measurements where made on the first peak of FP. Arrows indicate the time when bicuculline perfusion started into the MEA chamber. These data are representative of six slices.

glass electrodes. Bipolar versus monopolar current stimulations produce responses with a larger amplitude value because of more fibers recruitment as illustrated by FV measurements. Stimulation of Schaeffer collaterals will trigger action potentials that will travel along the axons until synaptic terminals in CA1. Such stimulus propagation, resulting from fast sodium channels opening, shall be fully inhibited by TTX (Debanne, 2004). We observe a strong inhibition of the FP with 500 nM TTX (Fig. 2) which is, however, not complete. The remaining “signal” observed after TTX inhibition shall correspond to a non-biological artefactual signal (see below), although TTX-independent glutamate release in CA1 has been reported (Strowbridge, 1999). In addition, we cannot totally rule out the influence of glial cells on the network excitability (Hama et al., 2004; Pascual et al., 2005). Because recorded signals are strongly depressed in low calcium–high magnesium medium, the FP constitute the sum of individual synaptic transmission steps occurring in the vicinity of the different recording electrodes. To assess that FP correspond to cell depolarizations mediated by AMPA/kainate receptors opening, we have

performed two types of experiments. First, recordings were performed with a half-sodium concentration. Second, we have used selective AMPA/kainate receptor antagonists. In a 50% decreased extracellular Na+ condition, the generation and the propagation of action potentials are not altered. However, and because of the modification of the electrochemical gradient, the amplitude of sodium-dependent depolarizations should be reduced by a half. AMPA/kainate receptors that mediate most of the fast glutamatergic transmission in the CA1 region are sodium-permeable ligand-gated channels (and also Ca2+-permeable for the ones not containing the GluR2 subunit); (Hollmann et al., 1991; Burnashev et al., 1992). Then, when the bath concentration is reduced by a half, excitatory post-synaptic currents should be reduced by 50%, and as a consequence, the postsynaptic membrane depolarization decreased by 50%. Indeed, in a halfsodium concentration of the bath, we observe a roughly 50% decrease of the FP amplitude. In addition, the FP is almost completely inhibited by saturating concentrations of CNQX or NBQX, confirming the preponderance of AMPA/kainate receptors in synaptic transmission in the CA1 region.

BR A IN RE S E A RCH 1 0 96 ( 20 0 6 ) 7 0 –8 4

79

Fig. 9 – Potentiation of GABAA receptors by neurosteroids could be detected by multi-electrode array recordings. (A) Bar chart presenting the mean increase of field potential amplitude (at steady-state effect) observed with different steroids known as selective GABAA receptors positive allosteric modulators: TH-PROG (100 nM; n = 26, 3 slices), TH-DOC (100 nM; n = 18, 2 slices), alphaxalone (3 μM; n = 17, 2 slices) and ganaxolone (1 μM; n = 18, 3 slices). The concentrations used correspond to the concentrations yielding maximal potentiation of currents evoked at GABAA receptors. One typical experiment performed with 1 μM ganaxolone, a synthetic and non-metabolized analog of allopregnanolone (TH-PROG), is detailed in panels (A), (B) and (C). (B) Localization of the stimulating and recording electrodes in the layers of the hippocampal slice. The stimulating electrode is indicated by the white filled circle. The recording electrodes are surrounded by a white line and traces presented in the other panels with respect to their geographical position within the slice. (C) Time course of the different FP amplitudes as a function of time. (D) Time course of slope factors as a function of time. In panels (C) and (D), the arrows indicate the time when ganaxolone was perfused on the slice. Error bars correspond to SEM. (***) Indicates that results are statistically significant (P < 0.001; one sample t test).

The cyanide-derivative FCCP fully depolarizes mitochondria and triggers fast energetic failure. Because mitochondria provide most, if not all, of the energy requested for the synaptic transmission machinery (Zenisek and Matthews, 2000; Li et al., 2004; Schuman and Chan, 2004), it would be expected that FCCP completely inhibit signals recorded in the slice. However, we still observe a slight remaining signal after FCCP poisoning and even when slices are oxygen-deprived for 20 min. Thus, the residual “apparent depolarization” might correspond to non-biological signals (artefact), and it is not observed using conventional glass recording techniques (Dominique Muller, personal communication). This residual signal is not dependent from the stimulating configuration, because it could be observed either with monopolar or

bipolar stimulations. Residual charges on the electrode tips and/or some capacitive coupling between the electrode tracks could be invoked to explain the residual signal that is observed on recording electrodes following TTX, AMPA/ kainate antagonists or FCCP exposures. Multiple forms of synaptic plasticity have been characterized in different brain areas and in the hippocampus. We illustrate here that some classical protocols employed to investigate some forms of short- or long-term plasticity could be run with a MEA. The performance of paired-pulse protocol is very easy and facilitation could be observed as long as the stimulus intensity is set just over the threshold to evoke FP responses. Such protocol has an interest when trying to investigate the effect of

80

BR A IN RE S EA RCH 1 0 96 ( 20 0 6 ) 7 0 –84

Fig. 10 – Ganaxolone potentiation of field potentials is observed simultaneously with a MEA and glass electrodes. (A) Comparison of the time courses of FP recorded simultaneously on both sides of the slice by a MEA electrode (on the bottom of the slice; panel A) and by a glass electrode (on the top of the slice; panel B). Traces recorded in control conditions and after reaching the steady-state effect of ganaxolone are superimposed. Data are representative of three independent experiments.

molecules on the presynaptic modulation of neurotransmitter release. Since its initial characterization (Bliss and Lomo, 1973), long-term potentiation (LTP) is one of the most widely used protocols to investigate the molecular basis of mid- to longterm memory (Collingridge and Bliss, 1995; Malenka and Nicoll, 1999). In the CA1 region, LTP is strongly dependent from the activation of the NMDA receptors (Davis et al., 1992; Collingridge and Bliss, 1995). As presented in Fig. 4, we observe that a selective competitive antagonist of NMDA receptors (D-AP5) completely blocks the induction of LTP as described in other studies (Davis et al., 1992; Hanse and Gustafsson, 1994; Wang and Stelzer, 1996; Jensen et al., 2003). This inhibition is fast and total only 10 min after DAP5 perfusion. We also observe a strong region-specific difference in the increase of FP following LTP induction; the LTP amplitude being much larger in the stratum pyramidale layer than in the stratum radiatum layer. Such regional differences have already been reported in other studies using standard glass electrodes (Kramar and Lynch, 2003; Vreugdenhil et al., 2005). This observation illustrates that quite refined regional investigations could be performed with the MEA recording technique. Long-term depression is another form of synaptic plasticity that was initially characterized in the cerebellum (Ito, 1989) using low-frequency stimulation protocols. Following this initial observation, it was shown that LTD could be induced in the CA1 region of the hippocampus when applying low-frequency stimulations (Stanton and Sejnowski, 1989; Muller et al., 1995). Activation of NMDA receptors is necessary for the induction of LTD (Mockett et al., 2002). More recently, it was revealed that another form of LTD could be induced in the CA1 region with the mGluR1/ mGluR5 selective agonist DHPG (Palmer et al., 1997; Bortolotto et al., 1999). This chemically induced LTD does not depend on NMDA receptor activation. However, the roles of post- or presynaptic mechanisms in this form of LTD as well as the signaling pathways implicated are the subject of

intense studies (Hou and Klann, 2004; Naie and ManahanVaughan, 2005). In our experiments, the induction of LTD by DHPG was very easy and reproducible. The amplitude of depression that we observe in standard physiological conditions (about 30%) is in the same amplitude range to what has been observed by conventional glass recording techniques (Fitzjohn et al., 2001). One of the most straightforward characteristic of MEA is to provide the ability of simultaneous recording of signals in different regions of brain slices and to observe differential effects of compounds according to the region of interest. We have chosen to illustrate this point by using gabazine, a selective GABAA antagonist that has the unique property to discriminate between GABAA receptors involved in fast synaptic transmission (phasic receptors) or extrasynaptic GABAA receptors mediating a tonic transmission (tonic receptors). Recent studies have described that micromolar or submicromolar concentrations of gabazine can selectively block the GABAA phasic current of hippocampal neurons (Bai et al., 2001; Stell and Mody, 2002; Yeung et al., 2003). In such conditions, the existence of a tonic GABAA current has been documented in granule cells of the dentate gyrus (Nusser and Mody, 2002) and in neurons of the CA1 region (Semyanov et al., 2003), but there is still a matter of debate on the precise localization of this current (pyramidal cells versus interneurons; (Farrant and Nusser, 2005). Thus, we find of interest to investigate what could be the effect of low or high gabazine concentrations in the CA1 region recorded with a MEA. As illustrated in Fig. 6, we observe that 500 nM gabazine slightly increase the FP recorded by the electrodes located close to the stratum pyramidale, whereas a 50 μM concentration of gabazine triggers a huge and selective increase of FP recorded close to the stratum pyramidale (see Fig. 7). However, epileptiform activities are observed on the electrodes located in stratum pyramidale and stratum radiatum as well. These effects that are very similar to the ones observed with 20 μM bicuculline (Fig. 8), in accordance with previous studies (Chapman et al., 1998), indicating that bicuculline

BR A IN RE S E A RCH 1 0 96 ( 20 0 6 ) 7 0 –8 4

inhibition modulates evoked responses in CA1 pyramidal neurons consecutively to Schaeffer collaterals stimulation. Bicuculline and gabazine induce oscillations together with an increase of evoked responses (due to balance excitation/ inhibition impairment), but these compounds never induce spontaneous epileptiform activities in the CA1 region of the hippocampus. It is of value to observe that most of pyramidal neurons inhibition seems to depend on the activity of tonic GABAA receptors (compare data of Figs. 6 and 7). As it has been recently described (Stell et al., 2003), we have recorded a slight potentiation of CA1 FP by the neurosteroids 3α, 5α-TH-PROG and 3α, 5α-TH-DOC. Both compounds are known as the most potent positive allosteric modulators of GABAA receptors. We observe that ganaxolone is more potent than 3α, 5α-TH-PROG and 3α, 5α-TH-DOC, very likely because it is not metabolized by the endogenous 3α-reductase (Belelli and Herd, 2003). Another synthetic steroid, alphaxalone, is less active, but this is coherent if the potentiation results from positive modulation of GABAA receptors. Namely, alphaxalone is ten times less potent than 3α, 5α-TH-PROG in modulating reconstituted GABAA receptors (Lambert, 1999). Nevertheless, the pharmacological signatures of steroids that increase the FP recorded in CA1 strongly suggest that such effect depends on the potentiation of GABAA receptors. If antagonists (gabazine or bicuculline) of GABAA receptors are able to increase FP, then it could be logically expected that positive allosteric modulators of GABAA receptors would have the opposite effect. However, we did observe the contrary. Subtle mechanisms of inhibition could explain such a result. Indeed, hippocampus interneurons form a complex and heterogeneous network and do not behave as “all or none” switches (Cossart et al., 2005). As recently discussed (Semyanov et al., 2004; Farrant and Nusser, 2005), tonic GABAA currents could modulate the excitability of GABAergic interneurons and their inhibitory drive on pyramidal cells. The authors propose that potentiation of GABAA tonic conductances could shunt interneurons excitability and, as a result, decrease the inhibition they exert on pyramidal cells. MEA recordings allow us to illustrate that positive modulation of tonic GABAA receptors by neurosteroids slightly increase the excitability of CA1 neurons, whereas inhibition of tonic and phasic GABAA receptors enhance their excitation in a much larger way. In conclusion, we believe that MEA recording is well suited for mid-throughput characterizations of compounds on brain slices. Such physiological and pharmacological profiling constitute “high content” steps in drug development and drug optimization processes.

and European legislations for animals care. The rats were sacrificed by fast decapitation, without previous anesthesia. The brain was quickly removed and soaked in ice-cold oxygenated buffer with the following composition in mM: KCl 2, NaH2PO4 1.2, MgCl2 7, CaCl2 0.5, NaHCO3 26, glucose 11, and saccharose 250. Hippocampus slices (300 μm) were cut with a MacIlwain tissue-chopper and incubated at room temperature for at least 1 h in Artificial Cerebro-Spinal Fluid (ACSF) of the following composition in mM: NaCl 126, KCl 3.5, NaH2PO4 1.2, MgCl2 1.3, CaCl2 2, NaHCO3 25, and glucose 11, bubbled with carbogen (95% O2, 5% CO2).

4.2.

Electrophysiological recordings and signal processing

4.2.1.

Positioning the slice on MEA

To cover the electrodes field (1.4 mm2), the slice is placed in the center of a 3D-MEA. Once the slice has settled down, the surrounding solution is fully removed (but not completely dried) with a tissue paper in order to ensure a tight adhesion between the slice and the electrodes. Then a netting ballast (U-shaped platinum wire with regularly spaced hair pieces) is carefully disposed on the slice to immobilized it. The MEA are quickly transferred to the amplifier stage on an inverted Olympus microscope, and the slice is continuously perfused with oxygenated ACSF (3 ml/min at 37 °C) during the whole recording session. The bath is ground connected with a 1 M NaCl-Agar bridge.

4.2.2.

Experimental procedure

4.1.

Preparation of acute hippocampal slices

Experiments were carried out with 17- to 21-day-old SpragueDawley rats (Elevage Janvier, Le Genest St Isle, France). Animals were housed and used in accordance to the French

Multi-electrode array set-up

All data were recorded with a MEA set-up from Multi Channel Systems MCS GmbH (Reutlingen, Germany) composed of a 4-channel stimulus generator and a 60-channel amplifier head-stage connected to a 60-channel A/D card. Software for stimulation, recordings and analysis were the ones commercially available from Multi Channel Systems: MC Stim (2.0.3.0 release) and MC Rack (3.2.1.0 release). All of the experiments were carried out with 3-dimensional MEA (Ayanda Biosystems, S.A., CH-1015 Lausanne, Switzerland) that consist of 60 tip-shaped and 60-μm-high electrodes spaced by 200 μm. The electrodes are made of platinum with 250 kΩ < impedance < 450 kΩ. Additional characteristics about these 3D-MEA have already been published (Heuschkel et al., 2002; Heuschkel et al., in press). MEA could be re-used approximately for about 50 recording sessions (mean duration of a recording session: ∼1 h). Electrode properties could be maintained constant with careful cleaning with deionized water following each recording session. In addition, wide spectrum proteases could be used (on a weekly basis) because electrodes are “coated” with proteins from hippocampal slices following multiple recording sessions.

4.2.3.

4.

81

Stimulation protocols

Following visual observation through a CCD camera connected to an inverted microscope, one of the electrodes identified in the CA3 field (at CA3/CA1 border) was disconnected from its amplifier and used to stimulate Schaeffer collaterals: when not specified, the standard stimulation protocol consisted in a monopolar biphasic current pulse (−300 μA 50 μs and +300 μA 50 μs) that was then injected

82

BR A IN RE S EA RCH 1 0 96 ( 20 0 6 ) 7 0 –84

through one of the selected electrodes every 30 s. When other stimulating protocols were performed (with current or voltage), they were biphasic and of identical duration (50 + 50 μs); only the current or voltage intensities were modified. Care was taken to choose the stimulating electrode in the same region from one slice to the other one.

4.2.4.

MEA recordings

Simultaneous evoked FP could usually be recorded from 6 and up to 12 electrodes located at the bottom of the slice in the CA1 stratum radiatum region. They correspond to the physiological propagation of the stimulus along the Schaeffer collateral pathway. FP from selected electrodes were simultaneously sampled at 5 kHz (or 10 kHz when necessary) and stored on the hard disk of a PC until offline analysis. For the recording and analysis of Fibers Volley, the sampling rate was increased up to 50 kHz. Without a slice in the chamber, the noise of the electrodes and set-up is in the range of 30–40 μV.

4.2.5.

Glass electrode recordings

In some experiments, the simultaneous recording of FP from the top of the slice was performed using a classical borosilicate glass electrode filled with the ACSF solution (resistance: 4–5 MΩ) and connected to one of the MEA headstage amplifiers with a fine isolated silver wire; this amplifier was physically disconnected from its MEA electrode by removal of the “contact pin”. The glass electrode was fitted into a Plexiglas electrode holder and was precisely positioned with a mechanical 3D micromanipulator. The glass electrode signal was processed like the other ones from MEA electrodes.

4.2.6.

LTP protocol

Long-term potentiation (LTP) was induced using a standard protocol (100 Hz LTP). Briefly, a tetanic stimulation (using the same current amplitude as the one employed to evoke FP) was applied for 1 s at 100 Hz and repeated twice with a 20-s interval. When necessary, FP recordings after LTP induction were performed for an additional 120-min period with one pulse stimulation every 30 s.

4.2.7.

Perfusion, drug applications and temperature control

During experiments, the slices were continuously perfused with ACSF solutions (bubbled with 95% O2–5% CO2) at the rate of 3 ml/min with a peristaltic pump (MEA chamber volume: ∼2 ml). Complete solution exchange in the MEA chamber was achieved 2 min after solutions exchange. Typically, FP were inhibited by 90% only 3 min after switching control saline for 50 μM CNQX. The perfusion liquid was continuously preheated at 37 °C just before to reach the MEA chamber with a heated perfusion cannula (Multichannel Systems, Reutlingen, Germany). The temperature of the MEA chamber was maintained at 37 °C with a Peltier element located in the MEA amplifier headstage.

4.2.8.

Offline analysis

FP amplitudes were measured as the difference between the baseline (before stimulation) and the maximal peak amplitude. For normalization, each FP amplitude was expressed as a percent of the mean-averaged amplitude recorded over a 10-

min control period, before drug application. Slope measurements were performed with the MC-Rack slope analysis module. Cursors were adjusted so that the slope calculation was performed on 90% of the total rise-time of the FP. Because electrodes are 200 μm spaced from each other, it could be estimated that the signals recorded at one electrode are independent from signals recorded at the neighboring ones. When necessary, the statistical effect of compounds was determined using one sample t test analysis.

4.3.

Chemicals

Compounds were purchased from Sigma-Aldrich (St Quentin Fallavier, France), TOCRIS (Avonmouth, United Kingdom) and Steraloids Inc. (Newport, USA) with the exception of ganaxolone that was custom-synthesized by Idealp SA (Lyon, France). Molecules to be tested were prepared at a 10 mM concentration in Dimethylsulfoxide (DMSO) (single use aliquots stored at −20 °C). Compounds were dissolved daily from stock-concentrated DMSO solutions into ACSF (bubbled with 95% O2–5% CO2); the final concentration of DMSO was ≤0.5%.

Acknowledgments The authors would like to thank Rebecca Pruss for her comments and suggestions all along this work, Valérie Crépel (U29, INSERM, INMED, Marseille) for her support and pieces of advice in the preparation of hippocampal slices and Pr. Dominique Muller (Dept. of Neuroscience, Faculty of Medicine, University of Geneva, Switzerland) and Michel Vignes (Montpellier II University, France) for the helpful criticisms and suggestions on preliminary data. We are indebted to Karl-Heinz Boven and the Multi Channel Systems' team for their continuous technical support.

REFERENCES

Bai, D., Zhu, G., Pennefather, P., Jackson, M.F., MacDonald, J.F., Orser, B.A., 2001. Distinct functional and pharmacological properties of tonic and quantal inhibitory postsynaptic currents mediated by gamma-aminobutyric acid(A) receptors in hippocampal neurons. Mol. Pharmacol. 59, 814–824. Belelli, D., Herd, M.B., 2003. The contraceptive agent Provera enhances GABA(A) receptor-mediated inhibitory neurotransmission in the rat hippocampus: evidence for endogenous neurosteroids? J. Neurosci. 23, 10013–10020. Belelli, D., Lambert, J.J., 2005. Neurosteroids: endogenous regulators of the GABA(A) receptor. Nat. Rev., Neurosci. 6, 565–575. Bianchi, M.T., MacDonald, R.L., 2004. Neurosteroid modulation of delta subunit containing GABA-A receptor channels. In: Smith, S.S. (Ed.), Neurosteroid Effects in the Central Nervous System. The Role of the GABA-A Receptor. CRC Press, Boca Raton, London, pp. 77–94. Bliss, T.V., Lomo, T., 1973. Long-lasting potentiation of synaptic transmission in the dentate area of the anaesthetized rabbit following stimulation of the perforant path. J. Physiol. 232, 331–356. Bortolotto, Z.A., Fitzjohn, S.M., Collingridge, G.L., 1999. Roles of

BR A IN RE S E A RCH 1 0 96 ( 20 0 6 ) 7 0 –8 4

metabotropic glutamate receptors in LTP and LTD in the hippocampus. Curr. Opin. Neurobiol. 9, 299–304. Burnashev, N., Monyer, H., Seeburg, P.H., Sakmann, B., 1992. Divalent ion permeability of AMPA receptor channels is dominated by the edited form of a single subunit. Neuron 8, 189–198. Chang, F.L., Greenough, W.T., 1984. Transient and enduring morphological correlates of synaptic activity and efficacy change in the rat hippocampal slice. Brain Res. 309, 35–46. Chapman, C.A., Perez, Y., Lacaille, J.C., 1998. Effects of GABA(A) inhibition on the expression of long-term potentiation in CA1 pyramidal cells are dependent on tetanization parameters. Hippocampus 8, 289–298. Collingridge, G.L., Bliss, T.V., 1995. Memories of NMDA receptors and LTP. Trends Neurosci. 18, 54–56. Cossart, R., Bernard, C., Ben-Ari, Y., 2005. Multiple facets of GABAergic neurons and synapses: multiple fates of GABA signalling in epilepsies. Trends Neurosci. 28, 108–115. Davis, S., Butcher, S.P., Morris, R.G., 1992. The NMDA receptor antagonist D-2-amino-5-phosphonopentanoate (D-AP5) impairs spatial learning and LTP in vivo at intracerebral concentrations comparable to those that block LTP in vitro. J. Neurosci. 12, 21–34. Debanne, D., 2004. Information processing in the axon. Nat. Rev., Neurosci. 5, 304–316. Doller, H.J., Weight, F.F., 1982. Perforant pathway activation of hippocampal CA1 stratum pyramidale neurons: electrophysiological evidence for a direct pathway. Brain Res. 237, 1–13. Farrant, M., Nusser, Z., 2005. Variations on an inhibitory theme: phasic and tonic activation of GABA(A) receptors. Nat. Rev., Neurosci. 6, 215–229. Fitzjohn, S.M., Palmer, M.J., May, J.E., Neeson, A., Morris, S.A., Collingridge, G.L., 2001. A characterisation of long-term depression induced by metabotropic glutamate receptor activation in the rat hippocampus in vitro. J. Physiol. 537, 421–430. Hama, H., Hara, C., Yamaguchi, K., Miyawaki, A., 2004. PKC signaling mediates global enhancement of excitatory synaptogenesis in neurons triggered by local contact with astrocytes. Neuron 41, 405–415. Hamill, O.P., Marty, A., Neher, E., Sakmann, B., Sigworth, F.J., 1981. Improved patch-clamp techniques for high-resolution current recording from cells and cell-free membrane patches. Pflugers Arch. 391, 85–100. Hanse, E., Gustafsson, B., 1994. TEA elicits two distinct potentiations of synaptic transmission in the CA1 region of the hippocampal slice. J. Neurosci. 14, 5028–5034. Heuschkel, M.O., Fejtl, M., Raggenbass, M., Bertrand, D., Renaud, P., 2002. A three-dimensional multi-electrode array for multi-site stimulation and recording in acute brain slices. J. Neurosci. Methods 114, 135–148. Heuschkel, M.O., Wirth, C., Steidl, E.-M. and Buisson, B., in press. Development of 3-D multi-electrodes arrays for use with acute tissue slices. In M. Baudry and M. Taketani (Eds.), Advance in Electrophysiology Using Multi-electrode Arrays, Kluwer Academic/Plenum, New York. Hodgkin, A.L., Huxley, A.F., 1952a. Movement of sodium and potassium ions during nervous activity. Cold Spring Harbor Symp. Quant. Biol. 17, 43–52. Hodgkin, A.L., Huxley, A.F., 1952b. A quantitative description of membrane current and its application to conduction and excitation in nerve. J. Physiol. 117, 500–544. Hollmann, M., Hartley, M., Heinemann, S., 1991. Ca2+ permeability of KA-AMPA-gated glutamate receptor channels depends on subunit composition. Science 252, 851–853. Hou, L., Klann, E., 2004. Activation of the phosphoinositide 3-kinase-Akt-mammalian target of rapamycin signaling pathway is required for metabotropic glutamate

83

receptor-dependent long-term depression. J. Neurosci. 24, 6352–6361. Hutzler, M., Fromherz, P., 2004. Silicon chip with capacitors and transistors for interfacing organotypic brain slice of rat hippocampus. Eur. J. Neurosci. 19, 2231–2238. Ito, M., 1989. Long-term depression. Annu. Rev. Neurosci. 12, 85–102. Jensen, V., Kaiser, K.M., Borchardt, T., Adelmann, G., Rozov, A., Burnashev, N., Brix, C., Frotscher, M., Andersen, P., Hvalby, O., Sakmann, B., Seeburg, P.H., Sprengel, R., 2003. A juvenile form of postsynaptic hippocampal long-term potentiation in mice deficient for the AMPA receptor subunit GluR-A. J. Physiol. 553, 843–856. Kramar, E.A., Lynch, G., 2003. Developmental and regional differences in the consolidation of long-term potentiation. Neuroscience 118, 387–398. Lambert, J.J., 1999. The selective interaction of neurosteroids with the GABA-A receptor. In: Baulieu, E.-E., Robel, P., Schumacher, M. (Eds.), Neurosteroids. A New Regulatory Function in the Nervous System. Humana Press, Totowa, pp. 125–142. Li, Z., Okamoto, K., Hayashi, Y., Sheng, M., 2004. The importance of dendritic mitochondria in the morphogenesis and plasticity of spines and synapses. Cell 119, 873–887. Majewska, M.D., Harrison, N.L., Schwartz, R.D., Barker, J.L., Paul, S.M., 1986. Steroid hormone metabolites are barbiturate-like modulators of the GABA receptor. Science 232, 1004–1007. Malenka, R.C., Nicoll, R.A., 1999. Long-term potentiation—A decade of progress? Science 285, 1870–1874. Meister, M., Wong, R.O., Baylor, D.A., Shatz, C.J., 1991. Synchronous bursts of action potentials in ganglion cells of the developing mammalian retina. Science 252, 939–943. Mockett, B., Coussens, C., Abraham, W.C., 2002. NMDA receptor-mediated metaplasticity during the induction of longterm depression by low-frequency stimulation. Eur. J. Neurosci. 15, 1819–1826. Muller, D., Hefft, S., Figurov, A., 1995. Heterosynaptic interactions between LTP and LTD in CA1 hippocampal slices. Neuron 14, 599–605. Naie, K., Manahan-Vaughan, D., 2005. Investigations of the protein synthesis dependency of mGluR-induced long-term depression in the dentate gyrus of freely moving rats. Neuropharmacology 49 (Suppl. 1), 35–44. Nusser, Z., Mody, I., 2002. Selective modulation of tonic and phasic inhibitions in dentate gyrus granule cells. J. Neurophysiol. 87, 2624–2628. Olsen, R.W., Cagetti, E., Wallner, M., 2004. Subunit specificity of steroid modulation of GABA-A receptors. In: Smith, S.S. (Ed.), Neurosteroid Effects in the Central Nervous System. The Role of the GABA-A Receptor. CRC Press, Boca Raton, London, pp. 47–61. Palmer, M.J., Irving, A.J., Seabrook, G.R., Jane, D.E., Collingridge, G.L., 1997. The group I mGlu receptor agonist DHPG induces a novel form of LTD in the CA1 region of the hippocampus. Neuropharmacology 36, 1517–1532. Pascual, O., Casper, K.B., Kubera, C., Zhang, J., Revilla-Sanchez, R., Sul, J.Y., Takano, H., Moss, S.J., McCarthy, K., Haydon, P.G., 2005. Astrocytic purinergic signaling coordinates synaptic networks. Science 310, 113–116. Schuman, E., Chan, D., 2004. Fueling synapses. Cell 119, 738–740. Semyanov, A., Walker, M.C., Kullmann, D.M., 2003. GABA uptake regulates cortical excitability via cell type-specific tonic inhibition. Nat. Neurosci. 6, 484–490. Semyanov, A., Walker, M.C., Kullmann, D.M., Silver, R.A., 2004. Tonically active GABA A receptors: modulating gain and maintaining the tone. Trends Neurosci. 27, 262–269. Shimono, K., Kubota, D., Brucher, F., Taketani, M., Lynch, G., 2002. Asymmetrical distribution of the Schaffer projections within

84

BR A IN RE S EA RCH 1 0 96 ( 20 0 6 ) 7 0 –84

the apical dendrites of hippocampal field CA1. Brain Res. 950, 279–287. Stanton, P.K., Sejnowski, T.J., 1989. Associative long-term depression in the hippocampus induced by Hebbian covariance. Nature 339, 215–218. Stell, B.M., Mody, I., 2002. Receptors with different affinities mediate phasic and tonic GABA(A) conductances in hippocampal neurons. J. Neurosci. 22, RC223. Stell, B.M., Brickley, S.G., Tang, C.Y., Farrant, M., Mody, I., 2003. Neuroactive steroids reduce neuronal excitability by selectively enhancing tonic inhibition mediated by delta subunit-containing GABAA receptors. Proc. Natl. Acad. Sci. U. S. A. 100, 14439–14444. Strowbridge, B.W., 1999. Glutamate receptors mediate TTX-resistant synchronous activity in the rat hippocampus. J. Neurosci. 19, 5758–5767. Tscherter, A., Heuschkel, M.O., Renaud, P., Streit, J., 2001. Spatiotemporal characterization of rhythmic activity in rat spinal cord slice cultures. Eur. J. Neurosci. 14, 179–190.

Vreugdenhil, M., Bracci, E., Jefferys, J.G., 2005. Layer-specific pyramidal cell oscillations evoked by tetanic stimulation in the rat hippocampal area CA1 in vitro and in vivo. J. Physiol. 562, 149–164. Wang, J.H., Stelzer, A., 1996. Shared calcium signaling pathways in the induction of long-term potentiation and synaptic disinhibition in CA1 pyramidal cell dendrites. J. Neurophysiol. 75, 1687–1702. Wirth, C., Luscher, H.R., 2004. Spatiotemporal evolution of excitation and inhibition in the rat barrel cortex investigated with multielectrode arrays. J. Neurophysiol. 91, 1635–1647. Yeung, J.Y., Canning, K.J., Zhu, G., Pennefather, P., MacDonald, J.F., Orser, B.A., 2003. Tonically activated GABAA receptors in hippocampal neurons are high-affinity, low-conductance sensors for extracellular GABA. Mol. Pharmacol. 63, 2–8. Zenisek, D., Matthews, G., 2000. The role of mitochondria in presynaptic calcium handling at a ribbon synapse. Neuron 25, 229–237.