| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
NEUROSCIENCE |
1 Toronto Western Research InstituteUniversity Health Network
2 Department of Medicine, (Division of Neurology)
3 Department of Physiology
4 Institute of Biomaterials and Biomedical Engineering
5 Epilepsy Research Program
6 Institute of Medical Science, University of Toronto, Toronto, Ontario, Canada M5T 2S8
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
(Received 7 June 2006;
accepted after revision 28 July 2006;
first published online 3 August 2006)
Corresponding author L. Zhang: Room 13-411, Toronto Western Hospital, 399 Bathurst Street, Toronto, Ontario, Canada M5T 2S8. Email: liangz{at}uhnres.utoronto.ca
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
These spontaneous population rhythmic activities have been shown to be involved in hippocampal synaptic plasticity and circuit development. Colgin et al. (2004) have explored the relationship between spontaneous field rhythms (
2 Hz) and long-term potentiation (LTP) in the CA1 area of conventional rat ventral hippocampal slices. Their data show that CA1 LTP is weak or absent in slices exhibiting the spontaneous field rhythm but can be uncovered after severing the Schaffer collateral projection and hence eliminating the CA1 spontaneous field rhythm. It is hypothesized that this spontaneous field rhythm constitutes an intrinsic mechanism that regulates memory-related synaptic plasticity. We have examined the postnatal development of the SRFPs in mouse hippocampal isolates (Wong et al. 2005). Our data show that SRFPs start to emerge in mouse hippocampal isolates around postnatal day 10, stabilize after postnatal day 15 and persist into adulthood. The postnatal development of SRFPs is in parallel with age-dependent modifications in the morphological and electrophysiological properties of rodent hippocampal neurons (cf. Wong et al. 2005), probably participating in activity-dependent consolidation processes of developing hippocampal networks. To demonstrate further the physiological significance of hippocampal SRFPs, in the present study we address the following issues: (1) whether the subiculum and entorhinal cortex (EC) exhibit similar SRFPs; (2) if so, what is the cellular basis of subicular and EC SRFPs; and (3) what is the temporal relation between hippocampal and subicular/EC SRFPs.
The subiculum and EC are the main output structures of the hippocampus. Subicular pyramidal neurons receive abundant axonal projections from CA1 pyramidal neurons, and the axons of subicular pyramidal neurons project into the EC and adjacent areas (Amaral & Witter, 1989; Lavenex & Amaral, 2000). The hippocampus, subiculum and EC function as an integrated system that is crucial for learning and memory processes (Naber et al. 2000; O'Mara et al. 2001; Battaglia et al. 2004; Squire et al. 2004; O'Mara, 2005), and abnormalities in this system have been implicated in pathological processes of Alzheimer's disease (Flood, 1991; Anderton et al. 1998), schizophrenia (Gray et al. 1991; Greene, 1996; Harrison, 2004) and temporal lobe epilepsy (Cohen et al. 2002; Wozny et al. 2003; Knopp et al. 2005). Whereas information is accumulating regarding the morphological and electrophysiological properties of individual neurons and pharmacologically induced rhythmic activities (Finch et al. 1983, 1988; Behr et al. 1996; Mason, 1993; Stewart & Wong, 1993; Greene & Totterdell, 1997; Colling et al. 1998; D'Antuono et al. 2001; Menendez de la Prida et al. 2003; Menendez de la Prida & Gal, 2004) in the subicular and EC areas, much less is known about the intrinsic network activities of the integrated hippocampal–subicular–EC circuits. It is therefore of interest to establish an in vitro model for further investigation of this issue.
We thus produced thick hippocampal–subicular–EC slices from adult mice and monitored regional rhythmic activities via extracellular and whole-cell patch recordings. Our data show that the SRFPs spread from CA3 towards the EC area and that the intrinsic rhythmicity of the subicular circuit plays a key role in spreading the SRFPs from the hippocampus to the EC. The present data have been presented previously in abstract form (Zhang et al. 2005).
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Slice preparation and in vitro perfusion
The animal was deeply anaesthetized with isoflurane and then decapitated. The brain was quickly removed, hemi-sectioned and immersed in an ice-cold, oxygenated artificial cerebrospinal fluid (ACSF) for 4–5 min before further dissection. The brainstem–thalamus tissues were then removed to expose the dorsal (or dentate gyrus) side of the hippocampus. Because the hippocampal fissure structurally divides the dentate gyrus and the CA2/CA1 areas, we applied a fine glass probe along the hippocampal fissure and gently separated the dentate gyrus from the CA2/CA1 areas while keeping dentate gyrus–CA3 connections. We then glued the brain tissue onto an aluminium block and obtained vibrotome slices along the hippocampal transverse plane. The thickness of slices was 1 mm for dorsal hippocampal–cortical sections and 0.7–0.8 mm for ventral sections because the CA3/CA1 striatum radiatum areas of the ventral hippocampus are wider (or thicker) than those in the dorsal hippocampus. After slicing, we separated hippocampal and adjacent cortical tissues along the lateral ventricle, and made a surgical cut near the rhinal fissure to produce S-shaped tissue pieces that encompassed the dentate gyrus, CA3, CA1 and subicular and EC areas (Fig. 1A and 1B). The thick hippocampal–EC slices were stabilized in an oxygenated (95% O2–5% CO2) ACSF at 32–33°C for 1–6 h before recordings. For recordings, the slice was placed in a submerged chamber and perfused with the oxygenated ACSF at 32–33°C. The perfusion rate was 15 ml min–1, and both top and bottom sides of the thick slice were effectively perfused in our recording chamber (see Wu et al. 2002 for a detailed description of our perfusion apparatus). The ACSF contained (mM): NaCl 125, KCl 3.5, NaH2PO4 1.25, NaHCO3 25, CaCl2 2, MgSO4 1.3 and glucose 10.
|
2 h before recordings in an attempt to minimize incision-associated disturbance of neuronal activities. In some experiments, we prepared conventional coronal slices (0.5 mm thickness, two slices per half brain) by vibrotome cuts from whole-brain chunks. These coronal slices contained ventral CA1/subicular and adjacent EC tissues but excluded CA2/CA3 areas (see Fig. 9A). These conventional slices were similarly maintained as described for the thick slices.
|
Glass electrodes (thin walled pipettes, 1.5 mm outside diameter, World Precision Instrument, Sarasota, FL, USA) were used for extracellular recordings. These electrodes were filled with a solution containing 150 mM NaCl and 2 mM HEPS (pH 7.4; resistance, 2 M
). Field potentials were recorded via a dual-channel amplifier (700A, Axon Instruments, Union City, CA, USA) and/or an extracellular recording amplifier (Model 300, AM Systems Inc., Carlsborg, WA, USA). The input frequency band of these amplifiers was set between DC and 3 kHz. Extracellular signals were further amplified (1000–5000 x) before digitization (Digidata 1200 or 1300, Axon Instruments).
Patch electrodes were pulled from the thin walled glass pipettes and filled with a solution containing (mM): potassium gluconate 140, KCl 10, Hepes 2 and EGTA 0.1 (pH 7.25). The resistance of patch electrodes was 4–5 M after being filled with the above solution. A ‘blind’ approach was used to search individual neurons in thick slices (Wu et al. 2005a,b).
O2-sensitive carbon fibre microelectrodes and a polarographic amplifier (Chemical Microsensor 1201, Diamond General, Ann Arbor, MI, USA) were used to measure tissue O2 levels in perfused slices (Wu et al. 2005a). These carbon fibre electrodes were calibrated in ACSF (at 32°C) that was aerated with 95% N2–5% CO2, 25% O2–5% CO2–70% N2, 50% O2–5% CO2–45% N2 or 95% O2–5% CO2. The output signals of the carbon fibre electrodes displayed a linear relation with the percentages of oxygen used for aerating the ACSF (r
0.95, linear regression analysis). As oxygen aeration was increased from 5% to 95%, the corresponding net increases in the O2 signals of these carbon fibre electrodes were in the range of 250–380 mmHg. The O2 levels measured from different depths of thick slices displayed an asymmetric ‘U’ shaped profile; that is, higher O2 levels at the top and bottom layers of the thick slices and the lowest O2 level in the middle layer of the slices. Such a ‘U’ shaped O2–depth profile is suggestive of effective perfusion of both top and bottom sides of the thick slice (see Wu et al. 2005a,b). The lowest O2 levels (middle layer) were 42.5 ± 10.3 mmHg for subicular areas and 66.0 ± 16.5 mmHg for EC areas (n
= 10 and 4 slices, respectively). These observations are in keeping with our previous measurements in the CA3/CA1 areas (Wu et al. 2005a,b), indicating adequate oxygenation of thick hippocampal–subicular–EC slices under our experimental conditions.
We used a bipolar tungsten wire electrode (diameter, 50 µm) for afferent stimulation. Constant current pulses (0.1–0.2 ms, 20–150 µA) were generated every 30 s by a Grass stimulator (S88) and delivered through an isolation unit. The stimulating electrode was placed in the dentate middle molecular layer, CA2 or subicular stratum radium area to stimulate the perforant pathway, Schaffer collateral pathway or subicular–EC projection pathway (arrowheads in Fig. 1A). For antidromic stimulation of subicular neurons, we placed the stimulating electrode in subicular alveus areas (circle in Fig. 1A). For EC stimulation, we placed the stimulating electrode in the medial EC area (square in Fig. 1A) in an attempt to activate a large number of medial EC cortical neurons. We routinely evoked CA1 synaptic field potentials as a functional indication of successfully prepared thick slices. The slices were excluded from the present data analyses if they did not exhibit stable synaptic field potentials or if their CA1 field EPSPs were 
0.5 mV following near maximal (100–150 µA) stimulation of the perforant pathway or the Schaffer collateral pathway.
Data analyses
Evoked synaptic field potentials and basic intracellular parameters were measured as previously described (Zhang et al. 1994, 1995, 1998; Wu et al. 2002, 2005a,b). To calculate the reversal potential of SRFP-correlated synaptic currents, pyramidal neurons were voltage-clamped at different voltages (–40 to –70 mV). The charge transfer of SRFP-correlated synaptic currents (pC) was measured from 30 consecutive events at each holding voltage. The mean values of these measurements were plotted against holding voltages and fitted with a linear regression function (r
0.90). The voltage at which the calculated linear regression line intersected the zero current level was taken as the reversal potential of SRFP-correlated synaptic currents. Spectral analyses were conducted using Micro-Origin (version 6, Origin Laboratory Corporation, Northampton, MA, USA) or pCLAMP software (version 9, Axon Instruments). To reduce the DC shift and high-frequency noise, original data were treated with a band-pass filter (0.2–0.5 Hz to 1000 Hz). Time–frequency analyses (Akay, 1997) of spontaneous field potentials were made from 65 s extracellular data segments, and the time resolution of these analyses was 0.5 s (Wu et al. 2005a). Statistical significance was determined using Student's t test or one-way ANOVA (SigmaStat or Origin). Mean and S.E.M. are presented throughout the text except where indicated.
Chemical and pharmacological agents
All solutions were made with deionized distilled water (specific resistance, 18.2 M cm–1; Milli-Q system). D-2-Amino-5-phosphopentanoate (AP5), 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) and bicuculline methiodide were purchased from Research Biochemicals Inc. (Canada).
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Simultaneous extracellular recordings from the CA1 and subicular areas revealed stable synaptic field potentials following single stimulation of the dentate gyrus area (Fig. 1C). These field potentials displayed a typical stimulation strength-dependent increase in amplitude (Fig. 1D), and the CA1 and subicular population spikes were 1.2 ± 0.1 and 0.3 ± 0.03 mV (n = 13 slices), respectively, in response to near maximal dentate gyrus stimulation (current pulses of 100–150 µA). EC synaptic field potentials were reliably induced by stimulation of the CA1 or subicular area (Fig. 1E). Following subicular stimulation, the amplitudes of subicular population spikes and EC deep layer field EPSPs were 0.80 ± 0.19 and 0.32 ± 0.08 mV (n = 16), respectively.
Intracellular activities of subicular and EC pyramidal neurons were examined via whole-cell patch recordings. The basic intracellular parameters of subicular pyramidal neurons were: resting potential, –59.5 ± 0.8 mV; input resistance, 84.3 ± 5.5 M; action potential amplitude, 95.4 ± 1.9 mV; and action potential half-width, 1.05 ± 0.05 ms (n
= 40). The intracellular parameters of EC deep layer (V–VI) pyramidal neurons were: resting potential, –63.6 ± 0.6 mV; input resistance, 148.5 ± 7.6 M; action potential amplitude, 96.9 ± 1.5 mV; and action potential half-width, 1.62 ± 0.14 ms (n
= 44). Five pyramidal neurons were recorded from EC superficial layers (I–II) with: resting potential, –66.5 ± 3.2 mV; input resistance, 76.8 ± 8.8 M; action potential amplitude, 95.3 ± 5.7 mV; and action potential half-width, 1.7 ± 0.37 ms. Generally, these measurements are in agreement with previous studies in conventional slices (Funahashi & Stewart, 1997; Mattia et al. 1997; Schmitz et al. 1998; Staff et al. 2000; Menendez de la Prida et al. 2003).
Previous studies have indicated that rat subicular pyramidal neurons exhibit two types of discharge patterns in response to intracellular stimulation: regular spiking and burst firing (Mattia et al. 1997; Staff et al. 2000; Menendez de la Prida et al. 2003). The burst firing is generally featured with a cluster of action potentials at frequencies of 200 Hz and is thought to result from high voltage-activated Ca2+ currents (Jung et al. 2001; but see Mattia et al. 1997). In our experiments, mouse subicular pyramidal neurons of thick slices also exhibited two types of discharge patterns following intracellular stimulation (depolarizing current pulses of 200–500 pA, 0.5–1 s). The burst firing was recognized by two to three spikes occurring in the first 200 ms of depolarizing pulses and the interspike intervals were 8 ms (Fig. 1F, middle trace). Of the 40 subicular pyramidal neurons recorded, 28 (70%) were considered as burst firing neurons, and the first and second interspike interval of the burst firing were 5.58 ± 0.31 and 6.90 ± 0.40 ms, respectively. The remaining 12 pyramidal neurons were considered as the regular spiking type (Fig. 1F, left-hand trace), and they discharged with evident spike adaptation but without early bursting firing. In response to intracellular injections of depolarizing current pulses, pyramidal neurons of EC deep and superficial layers discharged with evident spike adaptation but without bursting firing (Fig. 1F, right-hand trace).
To examine paired-pulse inhibition (Kapur et al. 1989) of the subicular circuit, we evoked antidromic subicular population spikes (see Methods) by identical twin stimuli with the interstimulus intervals varying from 5 to 50 ms (see Methods). As the twin stimuli were delivered at intervals of 5–10 ms, the second population spikes were greatly decreased as compared with the corresponding first population spikes (Fig. 1G). Similar paired-pulse inhibition was observed from individual subicular pyramidal neurons (n
= 5). One example is shown in Fig. 1H (top trace), where the first stimulus consistently induced action potentials with little variability in spike onset. The second stimulus, when applied 7 ms after the first stimulus, induced only a partial spike, which probably reflected an axon initial segment potential. When voltage-clamped at depolarizing potentials (–45 mV), this pyramidal neuron exhibited a large IPSC following a subthreshold or supra-threshold stimulus. The second supra-threshold stimulus, when applied near the peak of the evoked IPSC (6 ms after the first stimulus), was unable to evoke a spike current (Fig. 1H, bottom trace). In five of five subicular pyramidal neurons examined, the second supra-threshold stimulus failed to elicit an action potential or spike current when it was applied < 10 ms after the first stimulus. The IPSCs evoked by single subthreshold stimulation were 200–400 pA in amplitude (n
= 3 pyramidal neurons). These observations suggest that the local inhibitory circuit of subicular neurons (Menendez de la Prida, 2003) is functional in the thick hippocampal–subicular–EC slices, and that the subicular SRFPs presented below are not a consequence of disinhibition (Harris & Stewart, 2001a; Benini & Avoli, 2005).
Regional SRFPs
Multiple extracellular recordings revealed stable and coherent SRFPs in the CA3/CA1, subicular and EC areas of the thick slices. The frequencies of SRFPs varied within the range 0.5–4 Hz, with a mean frequency of 1.6 ± 0.2 Hz (n = 23 slices; Fig. 2A). The CA3/CA1 SRFPs displayed upward (positive) or downward (negative) waveforms when recorded from the striatum pyramidale or radiatum area, respectively. The waveforms of subicular SRFPs were largely downward, and upward SRFPs appeared only in the subicular alveus area. When recorded from CA and subicular alveus areas, the SRFPs were often superimposed with ripple-like oscillatory spike activities of 100–200 Hz. Because the appearance of such ripples was region-dependent and their intracellular correlates were not clearly identified (see also Wu et al. 2005b), we did not examine in detail the ripple-like oscillations in the present experiments.
|
As the onset times of SRFPs were often difficult to determine, we measured peak-to-peak time lags of regional SRFPs ( 60 consecutive events in each slice) in an attempt to explore their temporal relations. CA1 and subicular SRFPs exhibited a stable temporal relation and the peaks of CA1 SRFPs were time-advanced compared with corresponding subicular SRFP peaks in all slices examined (n
= 13, mean lags varying from 6.3 ± 0.5 to 29.3 ± 0.9 ms). The subicular and EC SRFPs appeared to occur simultaneously, and their peak-to-peak time lags could not be reliably measured because the EC SRFPs were small and their peaks were often difficult to recognize even after substantial filtering of the original data.
To explore the temporal relation of subicular and EC SRFPs, we monitored regional rhythmic activities before and after a surgical cut that separated the subicular and EC recording sites. Coherent SRFPs remained in the CA1 and subicular areas after the cut but regular rhythmic field potentials were absent in the isolated EC area (n = 4 slices; Fig. 2A). These observations could not be explained solely by non-specific damage to the EC circuit because the isolated EC areas were responsive to local afferent stimulation and showed spontaneous but irregular field potentials (0.05–0.2 Hz, data not shown). The latter observation is in keeping with recent in vitro studies of guinea-pig and rat EC neurons (Dickson et al. 2003; Kano et al. 2005). Thus, although the EC circuit is able to generate spontaneous population activities in isolation, the EC SRFPs observed from the ‘intact’ thick slices are largely dependent upon the subicular inputs.
To explore the possible influences of subicular neuronal activities on CA3 SRFPs, we surgically removed the CA1, subicular and EC areas and produced dentate gyrus–CA3 miniature slices (see Methods). In five dentate–CA3 miniature slices successfully prepared (CA3 population spikes of 0.96 ± 0.18 mV), stable SRFPs of 1.23 ± 0.17 Hz were observed from the CA3 area and their waveforms were comparable to those seen in the thick hippocampal–subicular–EC slices (see above). Thus, the activities of subicular neurons may have little influence on the CA3 SRFPs in the thick hippocampal–subicular–EC slices.
SRFP-correlated intracellular activities
Simultaneous extracellular and whole-cell patch recordings were used to examine SRFP-correlated intracellular activities. The exact distance between these two recording sites could not be determined under our experimental conditions. We estimate that these recordings sites were 300–600 µm apart.
Subicular pyramidal neurons.
When monitored at voltages near resting potentials (–60 mV), subicular pyramidal neurons exhibited periodic EPSPs or mixed EPSPs/IPSPs in correlation with local SRFPs and the peak-to-peak time lags between SRFPs and corresponding EPSPs varied in the range 3–25 ms. In the majority (27 out of 40) of subicular pyramidal neurons examined, the SRFP-correlated IPSP component became dominant when monitored at depolarizing voltages (–50 to –40 mV). As these neurons were depolarized to discharge, these periodic IPSPs effectively arrested tonic discharges (Fig. 3A).
|
Of the 40 pyramidal neurons examined, 13 neurons displayed only EPSPs/EPSCs in correlation with local SRFPs, even when monitored at –40 mV or more positive voltages (Fig. 3C). The lack of SRFP-correlated IPSP/IPSC components was unlikely to be due to a rundown of GABA-A receptor-mediated Cl– currents because such a phenomenon was recognizable in 2–3 min after forming the whole-cell recordings. Moreover, when monitored via cell-attached voltage-clamp recordings prior to whole-cell dialysis, these neurons often exhibited SRFP-correlated spike currents. The burst firing subicular pyramidal neurons appeared to have a high tendency of exhibiting SRFP-correlated EPSPs. In the 13 pyramidal neurons that exhibited only SRFP-correlated EPSPs/EPSCs, 11 neurons were of the burst firing type. These observations raise an interesting issue regarding the specific role of burst firing subicular pyramidal neurons in the generation of local SRFPs (see Discussion).
EC deep layer pyramidal neurons. SRFP-correlated synaptic activities were observed in 32 of 44 EC deep layer pyramidal neurons examined. When monitored near resting potentials (–60 to –65 mV), these neurons exhibited EPSPs/EPSCs in correlation with SRFPs (Fig. 4A). Of the 32 recordings, SRFP-correlated IPSP/IPSC components were noticeable in 17 neurons when monitored at depolarizing voltages (–50 to –40 mV). These IPSPs/IPSCs were relatively small as compared with preceding EPSPs/EPSCs (Fig. 4B). The reversal potentials of SRFP-correlated synaptic currents (see Methods) were estimated from seven pyramidal neurons that exhibited relatively large IPSCs at –50 and –40 mV (Fig. 4B). These estimated reversal potentials varied in a range from –57 to –40 mV, with a mean value of –50.1 ± 2.4 mV. Thus, Cl–-dominated currents are a major component of the SRFP-correlated synaptic activities in some EC deep layer pyramidal neurons.
|
Five pyramidal neurons were recorded from EC superficial layer (I–II) areas. When monitored near –70 mV, these neurons exhibited spontaneous EPSPs/EPSCs (Fig. 4D) but these synaptic activities were not correlated with the SRFPs simultaneously recorded from the EC deep layer or subicular area.
Observations made by dual whole cell recordings. To explore further the dynamics of SRFP-correlated synaptic activities, we performed dual whole-cell recordings together with extracellular monitoring of subicular SRFPs. In paired recordings made from proximal (near CA2) CA1 pyramidal neurons and subicular pyramidal neurons (n = 5 pairs), the SRFP-correlated IPSCs in the CA1 neurons led those in corresponding subicular neurons (Fig. 5C). As measured from 30 consecutive coherent IPSCs from each pair, the time lags in the onsets of CA1 and subicular IPSCs were 4.5 ± 0.8, 7.3 ± 0.4, 7.5 ± 0.3, 9.1 ± 0.8 and 12.8 ± 0.3 ms, (Fig. 5A). These observations are in keeping with the peak-to-peak measurements of CA1 and subicular SRFPs presented above and further demonstrate that in ‘intact’ slices the SRFPs spread from CA3 towards the subicular area.
|
1 µM bicuculline, which induced large amplitude rhythmic EPSPs or EPSPs/spikes in both subicular and EC pyramidal neurons. Under these conditions, the rising phases of EC EPSPs or EPSPs/spikes were delayed in relation to subicular EPSPs or EPSPs/spikes (Fig. 5D), suggesting a subicular-to-EC spread of excitatory activities in disinhibited thick slices. However, further experiments are needed to examine the temporal relation of SRFP-correlated synaptic activities between subicular and EC deep neurons.
Four paired whole-cell recordings were made from subicular pyramidal neurons. The SRFP-correlated EPSCs and/or IPSCs appeared to occur synchronously in these paired neurons without a clear time lag in their onsets (Fig. 5B). These observations suggest that subicular SRFPs are mediated by synchronized synaptic activities from a large population of pyramidal neurons. However, interneuronal synaptic connections were not evident in the paired neurons, as discharging one neuron by intracellular stimulation did not induce noticeable synaptic activities in the other corresponding neuron (Fig. 5E). Following perfusion of slices with 5–10 µM bicuculline, these paired subicular neurons exhibited synchronous epileptiform discharges but still failed to display interneuronal synaptic connections as tested by intracellularly induced discharges (data not shown). Previous studies have indicated that only closely neighboured pyramidal neurons ( 20 or 50 µm apart in the CA1 or EC area) are capable of exhibiting interneuronal synaptic connections when recorded simultaneously via microelectrodes (Deuchars & Thomson, 1996; Dhillo & Jones, 2000). Our dual whole-cell recordings were made from subicular pyramidal neurons 300 µm apart, which probably rendered a low probability of observing interneuronal synaptic connections in our experiments. Thus, further experiments to record from closely neighboured subicular neurons and to assess their morphological features in thick slices are necessary to characterize the subicular circuit responsible for the generation of SRFPs.
SRFPs originate from isolated subicular circuit
To explore the intrinsic rhythmicity of the subicular circuit, we adopted a miniature slice preparation as previously described by Harris & Stewart (2001b). We first prepared the thick slices and then surgically removed the dentate gyrus–CA3 and EC tissues (as indicated by dashed lines in Fig. 1A) while keeping CA1 and subicular areas. CA1–subicular synaptic connections were preserved in these CA1–subicular miniature slices. Following local CA1 stimulation at near maximal intensities, CA1 and subicular population spikes were 2.8 ± 0.3 and 0.6 ± 0.1 mV, respectively (n
= 25 slices). SRFPs were consistently observed from the subicular area of miniature slices. These SRFPs contained frequent, small-amplitude ( 30 µV) components as compared with those seen in the ‘intact’ thick slices (with CA3/EC inputs). Spectrogram analyses reveal that subicular SRFPs with or without CA3/EC inputs differ markedly in their frequency distribution profiles. As seen in Fig. 6, the subicular SRFPs without CA3/EC inputs encompass more persistent high-frequency (> 10 Hz) activities, whereas the subicular SRFPs with CA3/EC inputs contain periodical high-frequency activities which render significant variability in mean frequency over time (P < 0.05 over 10 s intervals). Such variability is absent in the subicular SRFPs without CA3/EC inputs.
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The rodent hippocampus exhibits two types of intermingled EEG activities during slow wave sleep or awake immobility: irregular activities and sharp waves (O'Keefe & Nadel, 1978; Buzsáki et al. 1983, 2003). The irregular activities have rhythmic signals over a wide frequency range (0.5–25 Hz) and their dominant frequencies are 2–3 Hz as revealed by spectral analyses (Leung et al. 1982; Jarosiewicz et al. 2002; Jarosiewicz & Skaggs, 2004a,b). The sharp waves have intermittent occurrence frequencies (0.01–0.5 Hz) and large amplitudes (up to 3 mV), and they are often superimposed with 200 Hz oscillatory spike activities called ripples (Buzsáki, 1986; Suzuki & Smith, 1987; Buzsáki et al. 2003). Hippocampal EEG sharp waves and irregular activities spread to the subiculum and EC deep layer areas (Chrobak & Buzsáki, 1994; Anderson & O'Mara, 2003), and the underlying network activities are thought to be crucial for hippocampus-dependent memory processes (Siapas & Wilson, 1998; Sirota et al. 2003; Jarosiewicz & Skaggs, 2004a,b; Battaglia et al. 2004).
The SRFPs we observed in vitro share some common features with the EEG irregular activities. Firstly, the in vitro SRFPs encompass a wide frequency range of rhythmic signals (0.5–20 Hz) and have dominant frequencies of 1–4 Hz (Wu et al. 2002, 2005a). Secondly, the SRFPs are often intermingled with an in vitro form of spontaneous sharp waves. The in vitro sharp waves are larger in amplitude (0.3–3 mV) and less frequent in occurrence (0.01–0.3 Hz) than the SRFPs. Distinct from the IPSP-based SRFPs, the in vitro sharp waves are correlated with coherent EPSPs/discharges in CA3 pyramidal neurons (Wu et al. 2005b). Thirdly, we show in the present experiments that the SRFPs can spread from CA3 to the subicular and EC areas (Fig. 3). Based on this information, we postulate that the network activities underlying SRFPs may participate in the generation of hippocampal EEG irregular activities.
Generation and regional spread of SRFPs
The CA3 recurrent circuit is known to play a prominent role in the generation of synchronous neuronal activities (Miles & Wong, 1986; Traub et al. 1989). The discharges of individual CA3 pyramidal neurons have been shown to effectively and preferentially excite local GABAergic inhibitory interneurons (Miles, 1990; Cohen & Miles, 2000; Fricker & Miles, 2000). We have shown in previous studies that the SRFPs originate from the CA3 area and that their intracellular correlates are characterized by synchronous GABA-A IPSPs in pyramidal neurons and repetitive discharges in inhibitory interneurons (Wu et al. 2002, 2005a). Based on this information, we have presented a hypothesis to account for SRFP generation and regional spread (Wu et al. 2005a). We propose that under standard in vitro conditions, sporadic discharges from a population of CA3 pyramidal neurons form a population excitatory drive which activates a group of GABAergic inhibitory interneurons. Once activated, the interneurons discharge rhythmically and coherently via mechanisms of mutual inhibition and gap junction couplings (Wang & Rinzel, 1993; Zhang et al. 1998; Skinner et al. 1999; Maier et al. 2003). The coherent interneuronal discharges then impose rhythmic IPSPs onto a large number of pyramidal neurons, which constitute a major ionic component of the SRFPs. We propose that the CA3 glutamate drive also excites CA1 interneurons via the Schaffer collateral pathway, thus mediating the local SRFPs in the CA1 area.
The SRFPs we observed from the isolated subicular circuit may be generated by similar mechanisms as proposed above. Harris et al. (2001) have shown that both regular spiking and burst firing pyramidal neurons of the rat subiculum have abundant interconnections, and the resulting excitatory activities can mediate synchronization and propagation of epileptiform discharges in the disinhibited subicular slices (Harris & Stewart, 2001a,b). Such pyramidal–pyramidal neuronal interconnections probably exist in the mouse subiculum. Moreover, because the ‘packing density’ of hippocampal neuronal elements is greater in mice than in rats (Insausti, 1993) and because the thick slices preserve a larger neuronal connectivity relative to the conventional slices, the interconnections of mouse subicular pyramidal neurons may form an excitatory drive which is sufficient to recruit local pyramidal neurons and inhibitory interneurons, and thereby to generate local SRFPs. The present observations that subicular SRFPs are blocked by pharmacological antagonists of GABA-A or AMPA glutamate receptors are supportive of this view. Considering that a great proportion (70% in our experiments) of subicular pyramidal neurons are burst firing neurons and that subicular SRFPs are correlated with mixed IPSPs/EPSPs in subicular pyramidal neurons, it is conceivable that the intrinsic network activities of the subicular circuit would have a high level of complexity. Such complexity may explain in part the persistent high-frequency (> 10 Hz) activities as revealed by spectrogram analyses of the SRFPs in CA1–subicular miniature slices (Fig. 6).
In ‘intact’ thick hippocampal–subicular–EC slices, although individually recorded CA1 pyramidal neurons barely show spontaneous discharges (Wu et al. 2002), it is possible that sparse discharges from a group of CA1 pyramidal neurons may be sufficient to influence the subicular networks and hence mediate the apparent spread of SRFPs from the CA1 to the subicular area. Previous morphological and electrophysiological studies have indicated that the axons of subicular pyramidal neurons project into the CA1 area (Berger et al. 1980; Köhler, 1985; Harris & Stewart, 2001b; Knopp et al. 2005). Such subicular–CA1 projections can induce excitatory responses in the CA1 pyramidal neurons (Commins et al. 2002; Shao & Dudek, 2005) and mediate the subicular–CA1 spread of epileptiform activities when examined in the presence of GABA-A receptor antagonists (Harris & Stewart, 2001a). It is conceivable that in the miniature CA1–subicular slices lacking CA3 inputs, the subicular–CA1 projections become dominant, thus spreading SRFPs from the subicular to the CA1 area.
Still, CA1 SRFPs can sometimes precede subicular SRFPs in the miniature CA1–subicular slices (Fig. 7A); the underlying mechanisms are unclear at present. One possibility is that the axon terminals of subicular pyramidal neurons synapse not only with CA1 pyramidal neurons but also with CA1 GABAergic inhibitory interneurons. As the CA1 SRFPs represent summed IPSPs from a large number of pyramidal neurons (Wu et al. 2002), the postulated innervation of CA1 interneurons by the subicular–CA1 projection, when effectively and preferentially activated (Miles, 1990; Cohen & Miles, 2000; Fricker & Miles, 2000), may promote the generation of CA1 SRFPs and hence partly account for the appearance of time advanced CA1 SRFPs in CA1–subicular miniature slices.
The mechanisms by which SRFPs are generated from the EC area remain to be examined. We show in the present experiments that a group of subicular pyramidal neurons, particularly the burst firing type, exhibit EPSPs/discharges in correlation with local SRFPs (Fig. 4), and that the EC SRFPs appear only in the deep layers which are the target of the axonal projection of subicular pyramidal neurons (Kloosterman et al. 2003). Moreover, the isolated EC area does not produce regular rhythmic activities (Figs 3 and 9). Based on these observations, we postulate that subicular inputs, especially the projection of burst firing subicular pyramidal neurons, are largely responsible for eliciting the EC SRFPs.
O'Mara et al. (2001) have postulated that the subiculum functions as an amplifier in hippocampal–cortical information transfer. In particular, subicular burst firing pyramidal neurons are thought to play an important role in the amplification of synaptic signals in both physiological and epileptiform activities. Our present data are supportive of this notion, suggesting that the subicular circuit serves as an amplifier to spread the SRFPs from the hippocampus proper to the EC. Specifically, we hypothesize that the discharges of subicular burst firing pyramidal neurons greatly promote excitatory activities in the subicular and EC circuits, probably via frequency-dependent facilitation or modulation of glutamate transmission (Silberberg et al. 2005), thus being crucial for the generation of subicular and EC SRFPs. To test this hypothesis, further experiments are needed to investigate how burst firing subicular pyramidal neurons interact among themselves and with other types of neurons in the subicular and EC areas.
Relevance of SRFPs to epileptiform activities
Schwartzkroin et al. (Schwartzkroin & Knowles, 1984; Schwartzkroin & Haglund, 1986) have examined spontaneous rhythmic activities of human temporal cortical neurons in brain slices that were obtained from surgical specimens of epileptic patients. Their data show that rhythmic IPSPs of 2–4 Hz occur synchronously in pyramidal neurons, and that these IPSPs are blocked by GABA-A receptor antagonists and are correlated with discharges of local inhibitory interneurons. Because similar synchronous rhythmic IPSPs are also observed in pyramidal neurons of normal monkey hippocampal slices, it is hypothesized that these rhythmic IPSPs are non-epileptogenic in nature and result from the activity of local GABAergic interneurons. Köhling et al. (1998) and Straub et al. (2000) have observed similar spontaneous rhythmic activities in slices of temporal lobe tissues of epileptic patients. Their data show that 40% of these slices exhibit spontaneous rhythmic field potentials of 0.5–1 Hz. These spontaneous field rhythms are correlated largely with Cl–-dependent IPSPs in local pyramidal neurons and are blocked by pharmacological antagonists of GABA-A or AMPA glutamate receptors.
Cohen et al. (2002) have examined subicular neuronal activities in brain slices prepared form surgical specimens of epileptic patients. Their data show that the subicular circuit can exhibit spontaneous rhythmic activities of 0.5–3 Hz, which are thought to result from network activities involving both pyramidal neurons and GABAergic inhibitory interneurons. Cohen et al. (2003) have hypothesized that these rhythmic activities represent a form of epileptogenic plasticity as a consequence of sclerotic loss of CA3/CA1 pyramidal neurons. Wozny et al. (2003) have observed similar spontaneous rhythmic activities in subicular neurons of human epileptic temporal lobe tissues but suggested that such subicular rhythmic activities are not necessarily related to CA3/CA1 sclerosis.
Knopp et al. (2005) have examined subicular neuronal activities in horizontal slices (0.4 mm thickness) prepared from sham control rats and rats with pilocarpine-induced chronic seizures. Their data show that spontaneous rhythmic field potentials of 0.5–2 Hz occur frequently in the subicular area of pilocarpine-treated rat brain slices but not in controls (see Fig. 2A of Knopp et al. 2005). It is hypothesized that seizure-induced strengthening of synaptic contacts within the subicular recurrent network is responsible for the generation of such population rhythmic activities.
We show in the present study that the subicular circuit of non-seizure mice is capable of exhibiting SRFPs of 1–4 Hz in vitro. The manifestation of subicular SRFPs in the thick slices may be partly due to a better preservation of neuronal circuits in thick slices than in conventional slices and partly due to the greater ‘packing density’ of neuronal elements in the mouse hippocampus (Insausti, 1993). Our data suggest that the subicular SRFPs are not a consequence of experimentally induced disinhibition. We hypothesize, rather, that the ability to generate SRFPs is an intrinsic property of the mouse subicular circuit, and that the cellular and network machineries underlying the subicular SRFPs are highly sensitive to modulation by chronic seizures (Cohen et al. 2002; Wozny et al. 2003; Knopp et al. 2005). It would be of significant interest in future studies to isolate a relatively large subicular circuit from animals with chronic temporal lobe epilepsy and to examine seizure-induced alterations in subicular SRFPs.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Amaral DG & Witter MP (1989). The three-dimensional organization of the hippocampal formation: a review of anatomical data. Neuroscience 31, 571–591.[CrossRef][Medline]
Anderson MI & O'Mara SM (2003). Analysis of recordings of single-unit firing and population activity in the dorsal subiculum of unrestrained, freely moving rats. J Neurophysiol 90, 655–665.
Anderton BH, Callahan L, Coleman P, Davies P, Flood D, Jicha GA, Ohm T & Weaver C (1998). Dendritic changes in Alzheimer's disease and factors that may underlie these changes. Prog Neurobiol 55, 595–609.[CrossRef][Medline]
Battaglia FP, Sutherland GR & McNaughton BL (2004). Hippocampal sharp wave bursts coincide with neocortical ‘up-state’ transitions. Learn Mem 11, 697–704.
Behr J, Empson RM, Schmitz D, Gloveli T & Heinemann U (1996). Electrophysiological properties of rat subicular neurons in vitro. Neurosci Lett 220, 41–44.[CrossRef][Medline]
Benini R & Avoli M (2005). Rat subicular networks gate hippocampal output activity in an in vitro model of limbic seizures. J Physiol 566, 885–900.
Berger TW, Swanson GW, Milner TA, Lynch GS & Thompson RF (1980). Reciprocal anatomical connections between hippocampus and subiculum in the rabbit evidence for subicular innervation of regio superior. Brain Res 183, 265–276.[CrossRef][Medline]
Buzsáki G (1986). Hippocampal sharp waves: their origin and significance. Brain Res 398, 242–252.[CrossRef][Medline]
Buzsáki G, Buhl DL, Harris KD, Csicsvari J, Czeh B & Morozov A (2003). Hippocampal network patterns of activity in the mouse. Neuroscience 116, 201–211.[CrossRef][Medline]
Buzsáki G, Leung LW & Vanderwolf CH (1983). Cellular bases of hippocampal EEG in the behaving rat. Brain Res 287, 139–171.[Medline]
Chrobak JJ & Buzsáki G (1994). Selective activation of deep layer (V-VI) retrohippocampal cortical neurons during hippocampal sharp waves in the behaving rat. J Neurosci 14, 6160–6170.[Abstract]
Cohen I & Miles R (2000). Contributions of intrinsic and synaptic activities to the generation of neuronal discharges in in vitro hippocampus. J Physiol 524, 485–502.
Cohen I, Navarro V, Clemenceau S, Baulac M & Miles R (2002). On the origin of interictal activity in human temporal lobe epilepsy in vitro. Science 298, 1418–1421.
Cohen I, Navarro V, Le Duigou C & Miles R (2003). Mesial temporal lobe epilepsy: a pathological replay of developmental mechanisms? Biol Cell 95, 329–333.[CrossRef][Medline]
Colgin LL, Kubota D, Jia Y, Rex CS & Lynch G (2004). Long-term potentiation is impaired in rat hippocampal slices that produce spontaneous sharp waves. J Physiol 558, 953–961.
Colling SB, Stanford IM, Traub RD & Jefferys JG (1998). Limbic gamma rhythms. I. Phase-locked oscillations in hippocampal CA1 and subiculum. J Neurophysiol 80, 155–161.
Commins S, Aggleton JP & O'Mara SM (2002). Physiological evidence for a possible projection from dorsal subiculum to hippocampal area CA1. Exp Brain Res 146, 155–160.[CrossRef][Medline]
D'Antuono M, Kawasaki H, Palmieri C & Avoli M (2001). Network and intrinsic contributions to carbachol-induced oscillations in the rat subiculum. J Neurophysiol 86, 1164–1178.
Deuchars J & Thomson AM (1996). CA1 pyramid-pyramid connections in rat hippocampus in vitro: dual intracellular recordings with biocytin filling. Neuroscience 74, 1009–1018.[Medline]
de Villers-Sidani E, Tahvildari B & Alonso A (2004). Synaptic activation patterns of the perirhinal-entorhinal inter-connections. Neuroscience 129, 255–265.[CrossRef][Medline]
Dhillon A & Jones RS (2000). Laminar differences in recurrent excitatory transmission in the rat entorhinal cortex in vitro. Neuroscience 99, 413–422.[CrossRef][Medline]
Dickson CT, Biella G & de Curtis M (2003). Slow periodic events and their transition to gamma oscillations in the entorhinal cortex of the isolated Guinea pig brain. J Neurophysiol 90, 39–46.
Finch DM, Nowlin NL & Babb TL (1983). Demonstration of axonal projections of neurons in the rat hippocampus and subiculum by intracellular injection of HRP. Brain Res 271, 201–216.[CrossRef][Medline]
Finch DM, Tan AM & Isokawa-Akesson M (1988). Feed forward inhibition of the rat entorhinal cortex and subicular complex. J Neurosci 8, 2213–2226.[Abstract]
Flood DG (1991). Region-specific stability of dendritic extent in normal human aging and regression in Alzheimer's disease. II. Subiculum. Brain Res 540, 83–95.[CrossRef][Medline]
Fricker D & Miles R (2000). EPSP amplification and the precision of spike timing in hippocampal neurons. Neuron 28, 559–569.[CrossRef][Medline]
Funahashi M & Stewart M (1997). Presubicular and parasubicular cortical neurons of the rat: functional separation of deep and superficial neurons in vitro. J Physiol 501, 387–403.[CrossRef][Medline]
Gnatkovsky V & de Curtis M (2006). Hippocampus-mediated activation of superficial and deep layer neurons in the medial entorhinal cortex of the isolated guinea pig brain. J Neurosci 26, 873–881.
Gray JA, Feldon J, Rawlins JNP, Hemsley DR & Smith AD (1991). The neuro-psychology of schizophrenia. Behav Brain Sci 14, 1–84.[Medline]
Greene JR (1996). Subiculum: a potential site of action for novel antipsychotic drugs? Mol Psychiatry 1, 380–387.[Medline]
Greene JR & Totterdell S (1997). Morphology and distribution of electrophysiologically defined classes of pyramidal and nonpyramidal neurons in rat ventral subiculum in vitro. J Comp Neurol 380, 395–408.
