| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Stanford Neuroscience Program and Department of Anesthesia, Neuropharmacology Laboratory, Stanford University School of Medicine, Stanford, California 94305-5117
 |
ABSTRACT |
|---|
 Abstract
 Introduction
Methods
Results
Discussion
References
|
|---|
Lukatch, Heath S. and M. Bruce MacIver. Physiology, pharmacology, and topography of cholinergic neocortical oscillations in vitro. J. Neurophysiol. 77: 2427-2445, 1997. Rat neocortical brain slices generated rhythmic extracellular field [microelectroencephalogram (micro-EEG)] oscillations at theta frequencies (3-12 Hz) when exposed to pharmacological conditions that mimicked endogenous ascending cholinergic and GABAergic inputs. Use of the specific receptor agonist and antagonist carbachol and bicuculline revealed that simultaneous muscarinic receptor activation and 
-aminobutyric acid-A (GABAA)-mediated disinhibition werenecessary to elicit neocortical oscillations. Rhythmic activity was independent of GABAB receptor activation, but required intact glutamatergic transmission, evidenced by blockade or disruption of oscillations by 6-cyano-7-nitroquinoxaline-2,3-dione and (±)-2-amino-5-phosphonovaleric acid, respectively. Multisite mapping studies showed that oscillations were localized to areas 29d and 18b (Oc2MM) and parts of areas 18a and 17. Peak oscillation amplitudes occurred in layer 2/3, and phase reversals were observed in layers 1 and 5. Current source density analysis revealed large-amplitude current sinks and sources in layers 2/3 and 5, respectively. An initial shift in peak inward current density from layer 1 to layer 2/3 indicated that two processes underlie an initial depolarization followed by oscillatory activity. Laminar transections localized oscillation-generating circuitry to superficial cortical layers and sharp-spike-generating circuitry to deep cortical layers. Whole cell recordings identified three distinct cell types based on response properties during rhythmic micro-EEG activity: oscillation-ON (theta-ON) and -OFF (theta-OFF) neurons, and transiently depolarizing glial cells. Theta-ON neurons displayed membrane potential oscillations that increased in amplitude with hyperpolarization (from 
30 to 90 mV). This, taken together with a glutamate antagonist-induced depression of rhythmic micro-EEG activity, indicated that cholinergically driven neocortical oscillations require excitatory synaptic transmission. We conclude that under the appropriate pharmacological conditions, neocortical brain slices were capable of producing localized theta frequency oscillations. Experiments examining oscillation physiology, pharmacology, and topography demonstrated that neocortical brain slice oscillations share many similarities with the in vivo and in vitro theta EEG activity recorded in other brain regions.
Rhythmic electroencephalographic (EEG) oscillations have been associated with both behaviorally relevant (Bland 1986 Slice preparation
Experiments were performed on brain slices isolated from juvenile male Sprague-Dawley rats (80-120 g) obtained from Simonsen Laboratories (Gilroy, CA). Experimental protocols were approved by the Institutional Animal Care Committee at Stanford University and adhered to published guidelines of the National Institutes of Health, Society for Neuroscience, and American Physiological Society. Rats were anesthetized with diethyl ether and their brains were removed into cold (1-2°C) oxygenated artificial cerebrospinal fluid (ACSF). The ACSF had the following ionic composition (in mM): 151.25 Na+, 3.5 K+, 2.0 Ca2+, 2.0 Mg2+, 130.5 Cl Micro-EEG recording
The term "micro-EEG activity" describes extracellular field activity recorded with the use of microelectrodes, which measure electrical voltage of a small, well-localized neuronal population. This nomenclature was adopted from that used by previous authors (Bullock and Basar 1988
Spectral analysis
Micro-EEG spectral quantification was accomplished with the use of fast Fourier transforms (FFTs) on 2.5-s epochs of data with the use of DataWave software. For each experiment, the frequency at peak power was extracted from averaged FFTs (n Temperature studies
An in-line Peltier temperature controller, provided by Dr. Keith Bley of Roche Bioscience (Palo Alto, CA), was used to control ACSF temperatures to an accuracy of 0.2°C. Temperature was stepped from 21 to 31°C in 0.3°C increments over 40 min (Fig. 3). Temperature was monitored in the perfusion chamber with the use of a calibrated miniprobe (Yellow Springs Instrument, Yellow Springs, OH) placed near (<5 mm) the brain slice.
Mapping studies
Intraslice electrode position visualization and hard copy output was accomplished with a Wild dissecting microscope, Optronics charge-coupled device camera, and Sony video printer, provided by Technical Instrument (San Francisco, CA). Neocortical laminar boundaries in Figs. 7 and 8 were based on histological measurements from two independent laboratories (Beaulieu 1993
Oscillation phase and amplitude analysis
A roving field recording electrode was used to collect micro-EEG data at various laminar depths along a trajectory normal to the pial surface (n = 5 slices). Interevent oscillation homogeneity was concurrently monitored in layer 2/3 with the use of a reference electrode positioned ~200 µm from the pial surface and <25 µm medial to the roving electrode trajectory. For phase analysis, reference electrode oscillation peak positivities were assigned a phase of 0°. Relative phase shifts were calculated by time aligning the reference and roving recordings, then dropping a vertical line from the peak positivity of the reference trace to intercept the roving trace (Fig. 7A). Roving oscillation phase at the point of interception was termed the relative phase shift. For example, if the peak positivities of both traces aligned perfectly (in phase), then a phase shift of 0° was assigned to that roving electrode position. If reference trace peak positivities aligned exactly with roving trace peak negativities, then a phase shift of 180° was assigned to that roving electrode position.
Current source density analysis and limitations
A DC-coupled roving micro-EEG recording electrode was moved perpendicularly from the pial surface to the underlying white matter in 15 consecutive 80-µm increments (16 recording sites), while a stationary recording electrode monitored response homogeneity in neocortical layer 2/3 (Fig. 8). Each response was initiated by stimulating (6 V, 500 µs, 0.033 Hz) the underlying white matter and deep layer 6 ~100 µm medioventral to the deepest recording site. One-dimensional current source density (CSD) profile estimates were constructed from laminar field potentials with the use of a finite difference formula (Mitzdorf 1985 Single-cell recording
Whole cell (>1-G Pharmacological agents
Carbamylcholine chloride (carbachol), (+)-tubocurarine chloride (curare), and atropine sulfate were obtained from Sigma (St. Louis, MO). ( Identification of neocortical micro-EEG oscillation generator
Neocortical brain slices perfused with carbachol (100 µM) and bicuculline (10 µM) exhibited trains of spontaneous rhythmic activity ranging in amplitude from 15 to 450 µV (Fig. 1, A and B). These oscillatory events began within 5-40 min of drug application, persisted for 1-4 h, appeared quasisinusoidal in nature, and were observed in both cortical hemispheres. To determine whether the oscillation generator was intrinsic to neocortex, a minislice of neocortex was dissected free (Fig. 1C) and exposed to carbachol (100 µM) and bicuculline (10 µM). The minislice displayed spontaneous rhythmic activity (Fig. 1D) similar both in form and frequency to the oscillations observed in intact brain slices. Averaging FFTs were performed on rhythmic activity, and the frequency at peak power was extracted for each slice (Fig. 1E). In each preparation, FFT analysis revealed a well-defined peak within the theta frequency range previously reported for rats (3-12 Hz). At room temperature, peak frequencies for 123 slices ranged from 3.8 to 9.5 Hz and followed a Gaussian distribution (
Oscillation temperature dependence
Oscillation peak frequency was temperature dependent, with a Q10 of 2.2 from 21 to 31°C. Within this temperature range, oscillation peak frequencies increased from 5 to 11 Hz, whereas oscillation amplitude and waveform appearance remained relatively constant (Fig. 3). As temperature increased, rhythmic epoch duration decreased (<3 s at 31°C). Total number of oscillations per epoch remained roughly constant at all temperatures examined. Thus increased oscillation frequencies observed at higher temperatures were associated with shortened oscillatory epochs.
Neocortical oscillation pharmacology
GABAergic.
Involvement of GABA in oscillatory activity was examined with the use of the GABAA receptor antagonist bicuculline, the chloride channel blocker picrotoxin, theGABAA receptor agonist muscimol, and the GABAB receptor antagonist CGP 35348. Bicuculline alone (10 µM, n = 15; 100 µM, n = 6; 500 µM, n = 5; or 1 mM, n = 3) elicited only large-amplitude (>150 µV) sharp spike activity in naive slices (Fig. 4A), similar to previously reported results in vivo (Mares et al. 1985
CHOLINERGIC.
To determine whether cholinergic receptor activation was sufficient to elicit rhythmic activity, the nonhydrolizable cholinergic agonist carbachol was used in isolation. Carbachol alone (100 µM, n = 4; 500 µM, n = 3; or 1 mM, n = 3) did not evoke oscillatory activity (Fig. 4D). However, because carbachol produced oscillatory activity when applied in combination with bicuculline, cholinergic receptor subtype involvement in neocortical oscillation generation was investigated with the use of the nicotinic receptor antagonist curare and the muscarinic receptor antagonist atropine. Similar to results obtained in vivo (Bland 1986 GLUTAMATERGIC.
Because the majority of neocortical synapses utilize glutamate as an excitatory neurotransmitter (Douglas and Martin 1990
Oscillation generator cortical topography and laminar distribution
MAPPING STUDIES.
The regional distribution of carbachol/bicuculline-induced neocortical oscillations was mapped with recordings from 51 slices across five consecutive coronal planes, each separated by 450 µm. On the basis of 249 recordings, oscillatory activity was localized to two bilaterally symmetric strips of cortex in the rostrocaudal plane (Fig. 6A). The medial-dorsal strip fanned out laterally at its posterior aspect, and, on the basis of previous histological measurements, included parts of four anatomically defined areas: 29d, 18b, 17, and 18a (Lysakowski et al. 1989
LAMINAR ANALYSIS.
Phase, amplitude, and CSD analysis provided detailed interlaminar information regarding oscillatory activity in area Oc2MM. Phase analysis (see METHODS) performed on 78 paired micro-EEG recordings in five slices revealed that oscillations reversed phase twice (Fig. 7). One phase reversal occurred 97 µm from the pial surface, whereas a second reversal was observed 916 µm deep, corresponding to layers 1 and 5, respectively (Fig. 7B). Layer 1 phase reversals were always more abrupt than layer 5 phase reversals, with 80% of the phase reversal occurring over 166 µm in layer 1 versus 358 µm in layer 5. Amplitude analysis demonstrated that large-amplitude oscillatory activity occurred primarily in superficial cortical layers, whereas attenuated signals were observed in deep layers (Fig. 7C). The largest normalized (see METHODS) oscillations (203 ± 76 µV, mean ± SD) were observed in layer 2/3, 255 ± 29 µm from the pial surface. In addition to oscillation phase and amplitude varying with cortical depth, oscillation waveform appearance also varied. Sinusoidal patterns were observed primarily in superficial layers, whereas middle layers exhibited sharper activity. Rhythmic activity with well-defined peaks separated by wide troughs was observed in deep layers and in the most superficial 60 µm of layer 1 (Fig. 7A).
Single-cell properties and theta frequency oscillations
OSCILLATION-ON (THETA-ON) NEURONS.
Cellular events underlying theta frequency micro-EEG oscillations were examined with the use of whole cell patch-clamp recordings in layers 2/3 and 5. Of 38 cells examined, three different cell types were identified on the basis of their response properties during micro-EEG oscillations. The most common cell type observed in both layer 2/3 (16 of 21 cells) and layer 5 (13 of 17 cells) was termed an "oscillation-ON" neuron. This neuronal subtype was characterized by the appearance of theta frequency membrane potential oscillations (MPOs, 2-15 mV) that rode on sustained depolarizations (4-25 mV) and were 180° out of phase with local oscillatory micro-EEG activity (Figs. 10A and 11A). In both layers 2/3 and 5, oscillation-ON neuron input resistances ranged from 100 to 400 M
GLIAL CELLS AND OSCILLATION-OFF (THETA-OFF) NEURONS.
The second most commonly encountered cell type in both layer 2/3 (5 of 21 cells) and layer 5 (3 of 17 cells) was glial cells. These cells had an average RMP of EXCITATORY CURRENTS APPEAR TO UNDERLIE OSCILLATION-ON (THETA-ON) NEURON ACTIVITY.
Intracellular membrane potential oscillations observed during rhythmic theta EEG activity in vivo have been associated with both excitatory postsynaptic potentials (EPSPs) (Fujita and Sato 1964 Unambiguous identification of an intrinsic neocortical theta generator in vivo has been hampered by volume conducted hippocampal theta EEG activity (Bringmann 1995 Neocortical micro-EEG oscillation similarities to theta EEG activity
Neocortical oscillation physiology and pharmacology were similar to previously described theta EEG activity in the following ways. 1) Neocortical slice oscillation frequencies and amplitudes (6.0 ± 1.1 Hz, 15-450 µV; Fig. 1) were comparable with previously reported in vivo theta EEG frequencies and amplitudes in rats (3-12 Hz, 50-1000 µV) for both limbic (Bland 1986 Oscillation similarities and differences with interictal epileptic EEG activity
Previous studies have demonstrated that depression of GABAA-mediated inhibition (Zia-Gharib and Webster 1991 Laminar specific oscillation characteristics
Because volume conducted EEG activity will not produce phase shifts or an increasing amplitude with increasing distance from a generator source (Holsheimer and Feenstra 1977 Neocortical oscillation single-cell properties
Previous studies in vivo have shown that activation of ascending cholinergic fibers results in sustained neocortical neuron depolarizations (Metherate and Ashe 1993 Synaptic mechanisms of neocortical oscillations
MPOs associated with theta activity in vivo have been ascribed primarily to either EPSPs (Fujita and Sato 1964 Oscillations and an ascending cholinergic synchronizing system
Cholinergically driven oscillations in neocortical brain slices augment a growing list of synchronized oscillatory activity currently under investigation. This list includes, but is not limited to, slow neocortical oscillations (Steriade et al. 1993b
INTRODUCTION
Abstract
Introduction
Methods
Results
Discussion
References
; Vanderwolf 1969) and pathological (Schaul 1990; Shinomiya et al. 1994; Steriade 1993) conditions. One well-studied class of oscillations is the 4- to 8-Hz theta EEG rhythm occurring in the hippocampus and associated limbic structures of rodents, cats, rabbits, and primates, including humans (Bland and Colom 1993). In freely moving rats, theta EEG activity, also termed rhythmic slow activity, encompasses a broader frequency range (3-12 Hz) (Lopes da Silva 1992; Whishaw and Vanderwolf 1973). The circuitry underlying hippocampal theta oscillations has yet to be fully elucidated; however, it has been shown that ascending medial septal cholinergic and GABAergic afferents are required for theta EEG generation (Colom et al. 1991; Smythe et al. 1992). Cholinergic inputs excite pyramidal neurons (Benardo and Prince 1982; Cole and Nicoll 1984; Halliwell and Adams 1982; Krnjevic 1993; Madison et al. 1987). Medial septal GABAergic inputs selectively innervate inhibitory interneurons (Freund and Antal 1988), suggesting that theta-associated activation of GABAergic afferents results in localized hippocampal disinhibition (Bilkey and Goddard 1985). Pharmacological or anatomic deafferentation of medial septal inputs from the hippocampus results in loss of theta rhythmicity. Theta-like activity can be restored with the use of agents that mimic medial septal influences. Two pharmacological agents that together have been shown to elicit robust theta frequency oscillations both in vivo (Colom et al. 1991) and in brain slices (Konopacki and Golebiewski 1993) are carbachol, a cholinergic agonist, and bicuculline, a -aminobutyric acid-A (GABAA) antagonist. Although pharmacologically induced theta frequency oscillations bear a striking resemblance to behaviorally relevant theta EEG activity (Konopacki 1996), their exact relationship with this activity remains unclear.
; Shute and Lewis 1967) and extrinsic GABAergic fibers that synapse preferentially onto local GABAergic interneurons (Freund and Gulyas 1991; Freund and Meskenaite 1992). This anatomy implies that neocortical circuitry may also be capable of generating oscillatory theta activity, as has been previously suggested (Landfield and McGaugh 1972; Yamaguchi et al. 1967). However, experimental findings in vivo have been controversial, with some groups observing an intrinsic neocortical theta generator within or near the posterior cingulate cortex (area 29) (Borst et al. 1987; Feenstra and Holsheimer 1979; Holsheimer 1982) whereas others have been unable to locate this generator (Colom et al. 1988; Gerbrandt et al. 1978). A confounding factor in these studies has been an inability to distinguish local neocortical theta EEG activity from volume conducted hippocampal theta activity (Bringmann 1995; Gerbrandt et al. 1978; Winson 1973). In the present study we used neocortical rat brain slices exposed to carbachol and bicuculline to determine whether intrinsic neocortical circuitry has the capacity to generate oscillations within the theta frequency range, and to further examine the cellular mechanisms underlying synchronized oscillatory activity within the neocortex.
METHODS
Abstract
Introduction
Methods
Results
Discussion
References
, 26 HCO3, 2.0 SO4, 1.25 H2PO4, and 10 glucose. Brains were sectioned in the coronal plane into 450-µm-thick slices with the use of a vibratome (Vibraslice Series 1000, Boston, MA). Before recording, slices were hemisected and placed on filter papers in a recovery chamber at the interface of a humidified carbogen (95% O2-5% CO2) gas phase and ACSF liquid phase. Slices from both hemispheres were allowed 
1 h to recover from the slicing procedure before submersion in ACSF in a recording chamber. The ACSF was saturated with carbogen gas and perfused at a rate of 2.0 ml/min at room temperature (21-24°C). Rapid and accurate solution changes were made with the use of a ValveBank8 computerized perfusion system (AutoMate Scientific, Oakland, CA).
; Petsche et al. 1984; Turbes 1993). In the present study, micro-EEG recording electrodes were low-resistance (<2 M
) glass microelectrodes filled with ACSF and placed in layer 2/3 of neocortex for most experiments (Fig. 1). Micro-EEG recording configuration was similar to that used for recording in vivo EEG activity: signals were amplified ×10,000-50,000 (model 210A, Brown-Lee Precision, San Jose, CA), filtered at 1-30 Hz band-pass, 60-Hz notch (Cyberamp 380, Axon Instruments, Foster City, CA), digitized at 512 or 2,048 Hz (DataWave Technologies, Longmont, CO), and stored on computer disk for further analysis.
View larger version (58K):
[in a new window]
FIG. 1.
Neocortical brain slices displayed oscillatory activity in the 3- to 12-Hz theta/rhythmic slow activity range when perfused with carbachol, a cholinergic agonist, and bicuculline, a -aminobutyric acid-A (GABAA) antagonist. A: coronal brain slice shows the neocortical region in which most recordings were made (between · · ·). In many experiments an extracellular microelectroencephalogram (micro-EEG) recording electrode (EEG) was placed in layer 2/3, whereas whole cell patch-clamp recordings (WC) were obtained either from layer 5 or 2/3. In some experiments oscillations were electrically evoked with a stimulating electrode (STIM). B: 2-s recordings show oscillatory activity in the presence of carbachol (100 µM) and bicuculline (10 µM). Scale bars: 0.2 s, 100 µV. C: minislice was excised from the neocortex and recordings were made from layer 2/3. D: carbachol and bicuculline elicited oscillatory activity from the minislice, demonstrating the presence of an intrinsic generator in this neocortical region. Scale bars as in B. E: plot showing the averaged fast Fourier transform (FFT) of the traces in B. In this experiment peak power occurred at a frequency of 5.6 Hz. F: histogram plot shows the distribution of peak power frequencies for 123 slices. The peak power frequency distribution in vitro (6.0 ± 1.1 Hz, mean ± SD) was within the frequency ranges reported for in vivo theta/rhythmic slow activity.
5) of carbachol/bicuculline-induced oscillatory activity.
View larger version (21K):
[in a new window]
FIG. 3.
Oscillation frequency was temperature dependent. A: micro-EEG traces obtained at progressively higher temperatures show that oscillation frequency increased with higher temperatures, whereas oscillation amplitude remained relatively constant. Scale bars: 100 µV, 0.2 s. B: oscillation frequencies (mean ± SD) are displayed at 0.3°C temperature increments between 21 and 31°C. The relationship between temperature and frequency was best fit with a 2nd-order polynomial curve that asymptoted near 12 Hz (- - -).
; Lysakowski et al. 1989).
View larger version (44K):
[in a new window]
FIG. 7.
Phase and amplitude analysis of paired electrode recordings revealed phase reversals across cortical layers 1 and 5, and peak oscillation amplitudes in layers 2/3. A: reference (ref) and stimulating (stim) electrode positions are shown in relation to hippocampal (HC) and cingulate cortices. Double lines: trajectory taken by the roving electrode during paired recordings. Time-aligned traces (2 s) show depth-dependent phase differences between reference and roving electrodes (· · ·). Trace amplitudes have been scaled to allow for phase comparisons. B: relative phase vs. cortical depth plot shows that oscillations reversed phase (90°) in layers 1 and 5. C: oscillation amplitude vs. cortical depth plot demonstrates that the largest-amplitude oscillations occurred in layer 2/3. In B and C, each point represents the mean ± SD for a 75-µm cortical depth binwidth; n = 5 slices.
View larger version (52K):
[in a new window]
FIG. 8.
Current source density (CSD) analysis of oscillatory activity revealed large current sinks in superficial cortical layers and current sources in layer 5. A: traces show DC-coupled roving micro-EEG recordings at 16 laminar depths and their respective stationary micro-EEG reference traces. CSD analysis was performed on DC-coupled recordings. An electrical stimulus ( 
with dashed lines) was used to trigger oscillations. Vertical dotted lines allow for assessment of stationary micro-EEG response homogeneity. B: 3-dimensional plot of CSD data shows pronounced current sinks in layer 2/3 and sources in deeper cortical layers. C: contour plot showing laminar location of current sources and sinks. Solid lines: sinks. Dashed lines: sources. Initial current flow occurred in deep cortical layers near the stimulating electrode position. Vertical lines: 1st, 3rd, 5th, and 7th maximal inward currents associated with oscillations. Note that peak current density location shifts from layer 1 to layer 2/3 by the 3rd oscillation. Each contour = 22.5 mV/mm2, with current densities of 0 ± 12.5 mV/mm2 omitted for clarity.
Where ROV is maximum peak-to-peak oscillation amplitude recorded with the roving electrode; REF is maximum peak-to-peak oscillation amplitude recorded with the reference electrode; and REFave is the average of maximum peak-to-peak oscillation amplitudes recorded with the reference electrode on the basis of 78 recordings, with one reference recording amplitude used per roving electrode position per slice (REFave = 198 ± 58 µV, mean ± SD).
; Nicholson and Freeman 1975). Cortical resistance along the axis was assumed to be constant (Mitzdorf and Singer 1980). Axum software (TriMetrix, Seattle, WA) was used for CSD calculations and for construction of three-dimensional surface and contour plots (Fig. 8).
; Buzsaki et al. 1986; Heynen and Bilkey 1994). Two-electrode data collection has also been used in neocortical brain slices to determine CSD profiles of epileptic (Hoffman et al. 1994) and nonepileptic (Cauller and Connors 1994) activity. In the present study, results from CSD analysis agreed well with results from both phase/amplitude analysis and laminar transection studies, suggesting that CSD error associated with response heterogeneity was minimal.
seal) patch recording microelectrodes (4-8 M) contained an internal solution composed of (in mM) 100 potassium gluconate, 10 ethylene glycol-bis(
-aminoethyl ether)-N,N,N
,N-tetraacetic acid, 5 MgCl2, 40 N-2-hydroxyethylpiperazine-N-2-ethanesulfonic acid free acid, 0.3 ATP, and 0.3 guanosine 5-triphosphate, pH 7.2, osmolarity 280-290 mosM. Whole cell recordings were obtained from layer 2/3 and 5 neurons and glia in Oc2MM neocortex. Signals were amplified ×50-1000 (Cyberamp 380, Axon Instruments), low-pass filtered at <10 kHz (Axon Instruments), digitized at 10 kHz (DataWave Technologies), and stored on computer disk for further analysis.
)-Bicuculline methiodide, picrotoxin, muscimol HBr, (±)-2-amino-5-phosphonovaleric acid (APV), and 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) were supplied by Research Biochemicals International (Natick, MA). CGP 35348 was provided by Ciba Geigy (Basel, Switzerland). All solutions were made up in spectrophotometric-grade water (Omnisolve) supplied by EM Science (Gibbstown, NJ). Chemicals for the ACSF and electrode solutions were reagent grade or better and were obtained from J. T. Baker (Philadelphia, PA).
RESULTS
Abstract
Introduction
Methods
Results
Discussion
References
2 = 36.5) with a mean peak frequency of 6.0 ± 1.1 Hz (Fig. 1F).
View larger version (37K):
[in a new window]
FIG. 2.
Multiple occurrences of carbachol/bicuculline-induced spontaneous and electrically evoked oscillatory activity are displayed on a slow time scale. A: in the presence of 100 µM carbachol and 10 µM bicuculline, spontaneous oscillatory epochs occurred 18 times in 10 min. B: under the same pharmacological conditions, trains of rhythmic oscillations were electrically evoked every 30 s by stimulation of deep cortical layer 6. B, top: AC-coupled micro-EEG recording. B, bottom: same data recorded in a DC-coupled configuration. Note that oscillatory epochs were associated with sustained extracellular negativities. Filled arrow: expanded rhythmic activity in C. Open arrow: failure of electrical stimulation to elicit oscillations. Failures occurred when spontaneous oscillations took place between electrical stimuli. C: expanded view showing that rhythmic micro-EEG oscillations ride on sustained extracellular negativities. A 2-s window (dashed box) represents the portion of an oscillatory epoch on which Fourier transform analysis was performed.
) and in vitro (Chagnac-Amitai and Connors 1989; Connors 1984). The ability of these slices to produce oscillatory activity was confirmed by adding carbachol to the perfusate, which resulted in the appearance of oscillatory activity in all slices within 5-40 min. Application of muscimol (5 µM, n = 5) to rhythmically oscillating slices blocked oscillations (Fig. 4B), indicating that enhanced tonicGABAA-mediated chloride flux and/or shunting inhibition can disrupt synchronous rhythmic activity.
View larger version (32K):
[in a new window]
FIG. 4.
GABAA receptor-mediated disinhibition and muscarinic receptor activation were necessary to elicit oscillations. A: control recordings show baseline micro-EEG activity in naive slices. The GABAA-receptor antagonist bicuculline (10 µM and 500 µM) evoked only large-amplitude nonrhythmic activity in slices capable of producing oscillations when carbachol was subsequently added. B: the GABAA agonist muscimol (5 µM) eliminated both oscillatory and nonoscillatory activity. C: rhythmic oscillatory activity was unaffected by either picrotoxin (100 µM), a GABAA chloride channel blocker, or CGP 35348 (100 µM), a GABAB receptor antagonist. D: the cholinergic agonist carbachol (100 µM and 1 mM) evoked no micro-EEG response when applied alone. E: carbachol/bicuculline-induced oscillations were not affected by the nicotinic receptor antagonist curare (50 µM), but were blocked by the muscarinic receptor antagonist atropine (1 µM).
; Zukin et al. 1974), it is possible that neocortical oscillations may require some intact inhibition. To determine whether GABAA-mediated inhibition was required for oscillatory activity, experiments were performed with the use of either the channel blocker picrotoxin (100 µM, n = 3; Fig. 4C) or high concentrations of bicuculline (100 µM, n = 3). Neither treatment had any effect on rhythmic activity, suggesting that these oscillations do not require GABAA-mediated inhibition.
). To test whether GABAB receptor activation was involved in oscillation generation, the GABAB-receptor antagonist CGP 35348 was used. CGP 35348 (100 µM, n = 5) had no effect on oscillations (Fig. 4C), suggesting that activation of GABAB receptors was not required for this rhythmic activity.
) and in vitro (Konopacki et al. 1988), curare (50 µM, n = 5) had no effect on carbachol/bicuculline-induced oscillations, whereas atropine (1 µM, n = 7) blocked oscillations, leaving only nonoscillatory activity resembling that observed with bicuculline alone (Fig. 4E). Thus simultaneous muscarinic receptor activation and GABAA disinhibition were necessary to evoke oscillatory micro-EEG activity in neocortical brain slices.
), glutamatergic involvement was investigated. Glutamate alone (50 µM, n = 4; 500 µM, n = 3) or in combination with bicuculline (10 µM, n = 4) did not elicit oscillatory activity (Fig. 5, A and B). However, intact glutamate-mediated transmission was necessary to maintain carbachol/bicuculline-induced oscillations, because CNQX (8.6 µM, n = 5), an 
-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA)/kainate receptor antagonist, reversibly blocked rhythmic activity (Fig. 5C), whereas APV (100 µM, n = 7), an N-methyl-D-aspartate (NMDA) receptor antagonist, disrupted oscillation timing, giving rise to unique semioscillatory activity (Fig. 5D). In four of seven experiments performed with APV, oscillatory activity similar to that shown in Fig. 5D was intermingled with larger-amplitude (>100 µV) bursting activity. In these cases, the most characteristic burst type was a spike doublet.
View larger version (37K):
[in a new window]
FIG. 5.
Glutamatergic transmission was necessary but not sufficient to evoke and maintain oscillations. A: application of L-glutamate (50 or 500 µM) did not elicit micro-EEG activity. B: glutamate (50 µM) in the presence of bicuculline (10 µM) elicited only nonoscillatory activity similar to that observed with bicuculline alone. C: the -amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA)/kainate glutamate receptor antagonist 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX, 8.6 µM) completely blocked carbachol/bicuculline-mediated oscillations. D: N-methyl-D-aspartate (NMDA) receptor antagonist (±)-2-amino-5-phosphonovaleric acid (APV, 100 µM) disrupted oscillation timing. E: nominally 0 mM magnesium elicited spontaneous slow oscillatory activity (note time scale) that was completely blocked by APV (100 µM). On washout of low magnesium and application of carbachol/bicuculline, spontaneous pharmacologically induced oscillations were observed.
), were dissimilar to carbachol/bicuculline-mediated activity. Low-Mg2+ oscillations progressively increased in amplitude, and were of low frequency (2.9 ± 0.6 Hz, mean ± SD; n = 6) and short duration (usually <3 s). In addition, low-Mg2+ oscillations were completely blocked by 100 µM APV (Fig. 5E), suggesting that these oscillations resulted from enhanced NMDA-mediated currents. Carbachol/bicuculline-induced oscillations were disrupted but not blocked by APV application (Fig. 5D), suggesting that additional (non-NMDA-mediated) mechanisms underlie carbachol/bicuculline-induced rhythmic activity.
), also known as the occipital association areas: Oc2MM, Oc2MM/Oc2ML, Oc1M/Oc1B, and Oc2L, respectively (Zilles 1985). The lateral strip included areas of piriform cortex and entorhinal cortex, the latter of which has been shown to generate theta EEG activity in vivo (Dickson et al. 1994; Mitchell and Ranck 1980) and in vitro (Konopacki et al. 1992b) under similar pharmacological conditions. Neocortical areas not displaying theta frequency oscillations in the presence of carbachol and bicuculline exhibited sharp spike activity similar to that observed with bicuculline alone, large-amplitude delta activity (Lukatch and MacIver 1996), or no activity at all (Fig. 6B).
View larger version (52K):
[in a new window]
FIG. 6.
Three-dimensional mapping studies revealed localized oscillatory micro-EEG activity within neocortex. A, right: 249 recordings in 51 slices map the presence ( ) or absence (
) of theta-like oscillations throughout a 2.25-mm-long volume of neocortex. Rhythmic oscillations were observed in 2 bilaterally symmetric strips of cortex that ran rostrocaudally. A, left: schematized rat brain shows the region from which slices were taken. B: representative traces from 3 slices show simultaneously recorded activity from Oc2MM, sensorimotor (SM), and entorhinal (EC) cortices. Note that micro-EEG activity in each cortical area was independent of activity in other areas.
View larger version (24K):
[in a new window]
FIG. 9.
Neuronal circuitry in superficial cortical layers was sufficient to generate theta frequency micro-EEG oscillations. Control micro-EEG recordings from layers 2/3 and 5 demonstrate synchronous oscillatory activity in these cortical layers. For comparison purposes, recordings from layer 5 were amplified ×5 more than layer 2/3 recordings. After transection of superficial and deep cortical layers, oscillatory micro-EEG activity was confined to layer 2/3, whereas layer 5 displayed only sharp spike nonoscillatory events similar to those observed with bicuculline alone.
. Oscillation-ON neurons could be either quiescent (24 of 29) or tonically active (5 of 29) at their resting membrane potential (RMP). For quiescent oscillation-ON neurons, average RMP was 60.1 ± 4.1 (SD) mV in layer 2/3 (n = 10) and 56.7 ± 4.3 mV in layer 5 (n = 14). RMPs of all tonically active oscillation-ON neurons were between 40 and 50 mV. In the absence of carbachol and bicuculline, RMPs ranged from 66 to 73 mV (mean = 70.9 ± 2.2 mV; n = 10), similar to previously reported RMPs of neocortical neurons in vitro (Jones and Baughman 1988).
View larger version (39K):
[in a new window]
FIG. 10.
Three cell types were observed during oscillatory micro-EEG activity. A: simultaneous micro-EEG and whole cell recordings show a tonically active "oscillation-ON" (theta-ON) neuron recorded in layer 5. During rhythmic micro-EEG oscillations this neuron depolarized and increased its firing rate. After oscillation cessation, a slight sustained membrane potential hyperpolarization (note - - -) was accompanied by decreased neuronal discharge activity. B: glia cell response [note resting membrane potential (RMP)], time-aligned to the oscillatory micro-EEG activity in A, exhibited a rapid depolarization during rhythmic micro-EEG activity, followed by a slow depolarization back to RMPs after oscillations ceased. C: unlike the oscillation-ON cell displayed in A, a time-aligned "oscillation-OFF" (theta-OFF) cell ceased tonic firing and became hyperpolarized (note - - -) during oscillatory micro-EEG activity. After termination of micro-EEG oscillations this neuron immediately returned to its preoscillation resting potential, unlike the cell types displayed in A and B.
View larger version (31K):
[in a new window]
FIG. 11.
Currents underlying oscillatory micro-EEG activity reversed polarity near 0 mV. A: simultaneous micro-EEG and whole cell current clamp recordings from a layer 2/3 oscillation-ON neuron demonstrated that membrane potential oscillation (MPO) amplitudes increased with hyperpolarization, whereas MPO frequency remained constant (- - -: RMP). Increased hyperpolarization also resulted in larger-amplitude, MPO-associated sustained depolarizing shifts. At depolarized potentials ( 37 mV), single-cell tonic discharge activity (
) did not register in micro-EEG recordings performed within 80 µm of the intracellular recording site, indicating that single-neuron discharge contributed relatively little to recorded micro-EEG activity. B: solid lines mark expanded micro-EEG activity and MPOs. A shorter time scale demonstrates that MPOs were ~180° out of phase with micro-EEG oscillations (37 mV). MPO peak positivities were frequently associated with undershooting action potentials, some of which are marked with arrows. At depolarized potentials (37 mV) only 1 in 5 oscillations evoked an action potential, perhaps owing to depolarization block of sodium channels at these potentials.
).
71.5 ± 4.5 mV, and input resistances of <100 M, consistent with previously reported passive membrane properties for glial cells (O'Connor et al. 1994). In addition, depolarizing current pulses never elicited action potentials in this cell type. Glial cell responses during oscillatory micro-EEG activity were characterized by sustained depolarizations (14.0 ± 3.5 mV, mean ± SD; Fig. 10B), which repolarized back to baseline RMP with a half-decay time of 4.8 ± 0.9 s after oscillation offset. Unlike oscillation-ON neurons, glia displayed either no MPOs or only small-amplitude MPOs (<0.2 mV).
).
; Nunez et al. 1987, 1990) and inhibitory postsynaptic potentials (IPSPs) (Leung and Yim 1986; Soltesz and Deschenes 1993; Ylinen et al. 1995). The nature of synaptic potentials underlying neocortical oscillation-ON neuron MPOs was investigated with the use of whole cell current injection. In 10 oscillation-ON neurons, membrane potentials were stepped from 30 to 90 mV, revealing maximal MPO amplitudes at the most hyperpolarized potentials and decreasing amplitudes with membrane depolarization (Fig. 11, A and B). Because neocortical EPSPs appear to reverse close to 0 mV (Sutor and Hablitz 1989), whereas chloride-mediated IPSPs reverse near resting potentials (Connors et al. 1988), this result suggests that EPSPs rather than IPSPs underlie MPO generation in oscillation-ON neurons. Although MPO amplitudes were quite sensitive to changes in membrane potential, MPO frequencies remained unchanged at all membrane potentials examined (Fig. 11). This voltage independence of MPO frequencies suggests that the primary drive for MPOs arises from network properties, rather than intrinsic properties of individual neurons (Ylinen et al. 1995).
DISCUSSION
Abstract
Introduction
Methods
Results
Discussion
References
; Gerbrandt et al. 1978; Winson 1973). The present study demonstrated that rat neocortical brain slices posses the requisite circuitry to generate synchronized oscillatory activity within the theta frequency range, consistent with earlier studies showing synchronized theta-like activity in brain slice preparations of hippocampus (Bland et al. 1988; Huerta and Lisman 1993; Konopacki et al. 1987; MacIver et al. 1986; MacVicar and Tse 1989) and entorhinal cortex (Konopacki et al. 1992b). It remains unclear, however, how representative in vitro neocortical oscillations are of theta EEG activity in vivo. To address this issue, in the present study we examined cellular and synaptic mechanisms underlying synchronous theta frequency neocortical oscillations.
; Bland and Colom 1993; Whishaw and Vanderwolf 1973) and neocortical structures (Holsheimer 1982). 2) Neocortical oscillation frequency temperature dependence (Q10 = 2.2; Fig. 3) was similar to that reported for hippocampal theta EEG activity in vivo (Q10 = 1.9) (Whishaw and Vanderwolf 1971). 3) Rhythmic micro-EEG oscillations were observed in the presence of carbachol (100 µM) and bicuculline (10 µM), two agents that together evoke robust theta EEG oscillations in vivo (Colom et al. 1991; Smythe et al. 1992) and in brain slices (Konopacki and Golebiewski 1993). 4) Neocortical oscillations required muscarinic receptor activation (Fig. 4), as did theta EEG activity recorded from the hippocampus in vivo (Bland 1986) and from brain slices of hippocampal (Konopacki et al. 1988; MacVicar and Tse 1989) and entorhinal cortex (Konopacki et al. 1992b). 5) Localization of neocortical oscillations to areas adjacent to posterior cingulate cortex (Fig. 6) matched that of previously reported neocortical theta EEG activity in vivo (Borst et al. 1987; Bringmann 1995; Feenstra and Holsheimer 1979; Holsheimer 1982), and these oscillations were not observed in other neocortical regions. 6) Neocortical slices, like hippocampal slices (MacVicar and Tse 1989), required intact glutamatergic transmission for theta frequency oscillations (Fig. 5), whereas depressed GABAA-mediated transmission did not alter rhythmic activity in hippocampal or neocortical slices (Fig. 4). 7) Oscillation-ON, oscillation-OFF, and glial cells described in the present study (Fig. 10) behaved similarly to theta-ON, theta-OFF (Konopacki et al. 1992a), and glial cells (Osehobo and Andrew 1993) characterized during theta activity in vivo and in vitro, respectively. 8) Theta EEG activity in vivo undergoes a well-characterized transition to delta, burst suppression, and isoelectric activity when exposed to increasing concentrations of barbiturate anesthetics (Buhrer et al. 1992; MacIver et al. 1996). Neocortical slice theta frequency micro-EEG oscillations displayed similar shifts in patterned activity when exposed to clinically relevant barbiturate concentrations (Lukatch and MacIver 1996).
) or enhancement of muscarinically mediated cholinergic transmission (Cruickshank et al. 1994) can lead to epileptic-like spiking activity. Use of 10 µM bicuculline in the present study may account for the finding that carbachol/bicuculline-induced oscillatory epochs occurred with a frequency (Fig. 2) similar to that previously reported for bicuculline-mediated interictal spikes (Avoli et al. 1995). In addition, the sharp spike micro-EEG negativity observed during neocortical oscillation commencement may represent the summated paroxysmal depolarizing shifts of a large neuronal population (Gutnick et al. 1982). It is possible that carbachol-induced muscarinic receptor activation gives rise to sustained plateau potentials by blocking rectifying K+ conductances (Fraser and MacVicar 1996; Krnjevic 1993; Madison et al. 1987), some of which have been suggested to terminate epileptiform burst discharges (Traub et al. 1994). Although this mechanism may contribute to prolonged epileptic bursts (Kriegstein et al. 1983), hippocampal theta EEG activity has also been shown to require muscarinic receptor activation and GABAA receptor suppression (Colom et al. 1991; Smythe et al. 1992). Thus it is not clear whether sustained rhythmic activity in neocortical brain slices should be thought of as a prolonged epileptic event, an abbreviated physiologically relevant oscillation, or some combination of both.
). Similarly, in the present study the muscarinic receptor antagonist atropine spared sharp spike discharge activity, but abolished all oscillatory activity (Fig. 4). 3) The NMDA receptor antagonist APV has been shown to block low-magnesium-induced seizure activity in this and previous studies (Zhang et al. 1994). APV disrupted, but did not block cholinergically induced neocortical oscillations (Fig. 5). 4) Previous studies have demonstrated that most, if not all, neocortical areas were susceptible to bicuculline-mediated seizure activity (Chagnac-Amitai and Connors 1989; Chervin et al. 1988; Sutor et al. 1994). In contrast, theta EEG activity was localized to specific brain regions (Bland 1986; Feenstra and Holsheimer 1979; Holsheimer 1982). In the present study, theta frequency oscillations were confined to a narrow localized region of neocortex (Fig. 6), which included those regions shown to support theta activity in vivo. 5) Laminar transection studies demonstrated that neuronal populations underlying oscillatory and sharp spiking activity were anatomically separable (Fig. 9). and 6) Previous reports have demonstrated that virtually all neurons display large-amplitude (>25 mV) paroxysmal depolarizing shifts during epileptic activity (Gutnick et al. 1982). Neuronal membrane responses during neocortical cholinergic oscillations were varied; some neurons displayed moderate depolarizations (~20 mV), whereas others hyperpolarized or displayed small membrane potential depolarizations (<5 mV) and low-amplitude MPOs (Figs. 10 and 11).
), in vivo studies seeking to identify intrinsic theta generators in cortex have constructed oscillation amplitude and phase depth profiles (Feenstra and Holsheimer 1979; Heynen and Bilkey 1994; Ylinen et al. 1995). Criteria for a local generator in these studies has been based on finding a phase reversal and/or an amplitude maximum within the cortical region of interest. In the present study we adopted the same criteria for localizing neuronal generators. It was first determined that neurons and synchronizing circuits generating theta frequency oscillations must be intrinsic to neocortex because isolated minislices of Oc2MM cortex produced oscillatory activity (Fig. 1). Within this neocortical area, a phase reversal was observed in layer 5 (Fig. 7B), consistent with previously reported theta frequency oscillation phase reversals in layer 5 in vivo (Feenstra and Holsheimer 1979; Holsheimer 1982). Layer 5 phase reversals in vivo and in vitro both occurred over ~350 µm. Neocortical brain slices also displayed a phase reversal in layer 1 that was not observed in vivo. In vivo studies likely did not detect this phase reversal because it occurred abruptly, and very close to the pial surface (<100 µm), at a depth not reported in vivo (Feenstra and Holsheimer 1979). Consistent with in vivo findings, neocortical brain slice oscillation phase reversals were associated with decreased oscillation amplitudes.
). This initial CSD profile may reflect a bicuculline-mediated epileptic-like event that over time (~250 ms) is replaced in layer 2/3 by muscarinic-mediated synchronous oscillations resembling theta EEG activity. Similarly, previous studies have demonstrated that separate mechanisms underlie synchronized neuronal discharges and depolarizing shifts associated with hippocampal electrographic seizure activity (Swann et al. 1993). Further support for the idea that two separate processes underlie neocortical micro-EEG oscillations came from laminar transection experiments that demonstrated that superficial cortical layers were capable of supporting sustained oscillatory activity, whereas isolated deep cortical layers produced only sharp spike activity (Fig. 9). The observance of epileptic-like activity in deep layers is consistent with previous studies that showed that the lowest threshold for epileptiform discharge occurred in neocortical layers 4 and 5 (Connors 1984; Prince and Connors 1986). However, to our knowledge, localization of theta frequency micro-EEG oscillations to superficial neocortical layers is a novel finding.
; Metherate et al. 1992). In the present study, similar intracellular depolarizations (4-25 mV) were observed during oscillatory activity (Figs. 10 and 11). These sustained depolarizations were accompanied by rhythmic MPOs that were 180° out of phase with local micro-EEG oscillations. Comparable MPOs have been observed during hippocampal theta activity, in vivo and in vitro, in which pyramidal neurons became depolarized by 10-20 mV and generated spontaneous MPOs of 5-25 mV, which were also 180° out of phase with respect to local EEG signals (Bland et al. 1988; Konopacki et al. 1992a; Nunez et al. 1987). Neurons involved in interictal spiking activity also demonstrate sustained depolarizing shifts, sometimes accompanied by MPOs (Hoffman et al. 1994). However, seizure-associated depolarizing shifts are usually significantly larger in amplitude (typically >30 mV), and MPOs are faster in frequency (typically >20 Hz), than those observed in the present study and those previously reported for theta activity in vivo. In addition, one hallmark of epileptogenesis, the simultaneous occurrence of large-amplitude paroxysmal depolarizing shifts in virtually every neuron, was not observed during oscillatory micro-EEG activity in neocortex; many oscillation-ON neurons displayed only small depolarizing shifts (<10 mV), and some neurons hyperpolarized during rhythmic oscillations (Fig. 10). Thus the behavior of these neurons more closely resembled previously reported theta-ON and theta-OFF cell activity displayed by neurons during physiologically occurring theta EEG activity (Bland and Colom 1993).
; Nunez et al. 1987, 1990) or IPSPs (Leung and Yim 1986; Soltesz and Deschenes 1993; Ylinen et al. 1995), and the exact nature of these cellular oscillations remains a matter of controversy. In the present study, neocortical micro-EEG oscillations were unaffected by GABAA and GABAB receptor antagonists, but were blocked by the glutamate receptor antagonist CNQX. This, together with the finding that oscillations decreased in amplitude as RMPs approached 0 mV, suggests that excitatory rather than inhibitory currents underlie these particular oscillations. Synchronized theta frequency oscillations were localized to superficial cortical layers, and peak oscillation amplitudes were observed in layer 2/3, suggesting that layer 2/3 neurons generated muscarinic-mediated rhythmic activity. Previous studies have demonstrated a population of cholinergic neurons intrinsic to layer 2/3 that may account for up to 30% of neocortical cholinergic fibers (Houser et al. 1985), and it is possible that these neurons may be involved in neocortical theta frequency oscillations in vivo. In addition, small groups of layer 2/3 neurons have been shown to be electrotonically coupled (Gutnick and Prince 1981), and this coupling could help sustain excitatory network reverberations that contribute to synchronous oscillation generation.
), thalamocortical oscillations (Steriade et al. 1993a), spindle activity in lateral geniculate slices (Bal et al. 1995; Kim et al. 1995), metabotropic receptor-mediated oscillations in hippocampal slices (Taylor et al. 1995), NMDA receptor-mediated rhythmic activity in slices of caudal nucleus tractus solitarii (Tell and Jean 1993), and rhythmic slow activity (theta) in hippocampal brain slices (MacIver et al. 1986; MacVicar and Tse 1989). Results from the present study fit a pharmacological and physiological profile of an ascending cholinergic synchronizing pathway that originates in the brain stem (Vertes et al. 1993) and projects via the hypothalamus to medial septum (Bland et al. 1994), hippocampus (Oddie et al. 1994), and midline neocortical structures (Feenstra and Holsheimer 1979; Holsheimer 1982), suggesting that many if not all limbic structures support theta EEG oscillations. Unlike the hippocampus, which receives its primary cholinergic afferents from the medial septum (Gaykema et al. 1990), medial neocortical areas generating theta frequency micro-EEG oscillations in the present study receive their primary cholinergic innervation from the magnocellular basal nucleus (Luiten et al. 1987). Although the function of cholinergically mediated oscillations remains to be determined, there is growing evidence that theta frequency synchronizations are involved in arousal, sensory/motor processing, learning, and memory (Bland and Colom 1993; Huerta and Lisman 1993).
| |
ACKNOWLEDGEMENTS |
|---|
We thank Drs. D. V. Madison and J. Huguenard for helpful comments on this study, F. Monroe for technical assistance and help with manuscript preparation, and G. Rabinovici for assistance with laminar transectionstudies.
This work was supported by National Institute of General Medical Sciences Grant GM-49811 to M. B. MacIver and a United States Air Force OSR/SCEEE graduate fellowship to H. S. Lukatch.
| |
FOOTNOTES |
|---|
Address reprint requests to M. B. MacIver.
Received 22 October 1996; accepted in final form 30 December 1996.
| |
REFERENCES |
|---|
|
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
This article has been cited by other articles:
|
|
 |
|
  R. Kirov, C. Weiss, H. R. Siebner, J. Born, and L. Marshall Slow oscillation electrical brain stimulation during waking promotes EEG theta activity and memory encoding PNAS, September 8, 2009; 106(36): 15460 - 15465. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. H. Higgs and W. J. Spain Conditional Bursting Enhances Resonant Firing in Neocortical Layer 2-3 Pyramidal Neurons J. Neurosci., February 4, 2009; 29(5): 1285 - 1299. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. E. Fanselow, K. A. Richardson, and B. W. Connors Selective, State-Dependent Activation of Somatostatin-Expressing Inhibitory Interneurons in Mouse Neocortex J Neurophysiol, November 1, 2008; 100(5): 2640 - 2652. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Bollimunta, Y. Chen, C. E. Schroeder, and M. Ding Neuronal Mechanisms of Cortical Alpha Oscillations in Awake-Behaving Macaques J. Neurosci., October 1, 2008; 28(40): 9976 - 9988. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J.-Y. Wu, Xiaoying Huang, and Chuan Zhang Propagating Waves of Activity in the Neocortex: What They Are, What They Do Neuroscientist, October 1, 2008; 14(5): 487 - 502. [Abstract] [PDF]
|
|
|
|
|
|
L. Bai, X. Huang, Q. Yang, and J.-Y. Wu Spatiotemporal Patterns of an Evoked Network Oscillation in Neocortical Slices: Coupled Local Oscillators J Neurophysiol, November 1, 2006; 96(5): 2528 - 2538. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. E. Fitch, R. N. Sahr, B. J. Eastwood, F. C. Zhou, and C. R. Yang Dopamine D1/5 Receptor Modulation of Firing Rate and Bidirectional Theta Burst Firing in Medial Septal/Vertical Limb of Diagonal Band Neurons In Vivo J Neurophysiol, May 1, 2006; 95(5): 2808 - 2820. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. S. Lukatch, C. E. Kiddoo, and M. B. MacIver Anesthetic-induced Burst Suppression EEG Activity Requires Glutamate-mediated Excitatory Synaptic Transmission Cereb Cortex, September 1, 2005; 15(9): 1322 - 1331. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
X. Huang, W. C. Troy, Q. Yang, H. Ma, C. R. Laing, S. J. Schiff, and J.-Y. Wu Spiral Waves in Disinhibited Mammalian Neocortex J. Neurosci., November 3, 2004; 24(44): 9897 - 9902. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. T. Dypvik and B. H. Bland Functional Connectivity Between the Red Nucleus and the Hippocampus Supports the Role of Hippocampal Formation in Sensorimotor Integration J Neurophysiol, October 1, 2004; 92(4): 2040 - 2050. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Bao and J.-Y. Wu Propagating Wave and Irregular Dynamics: Spatiotemporal Patterns of Cholinergic Theta Oscillations in Neocortex In Vitro J Neurophysiol, July 1, 2003; 90(1): 333 - 341. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. B. Caplan, J. R. Madsen, A. Schulze-Bonhage, R. Aschenbrenner-Scheibe, E. L. Newman, and M. J. Kahana Human {theta} Oscillations Related to Sensorimotor Integration and Spatial Learning J. Neurosci., June 1, 2003; 23(11): 4726 - 4736. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. H. Bland, J. Konopacki, and R. H. Dyck Relationship Between Membrane Potential Oscillations and Rhythmic Discharges in Identified Hippocampal Theta-Related Cells J Neurophysiol, December 1, 2002; 88(6): 3046 - 3066. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. D'Antuono, H. Kawasaki, C. Palmieri, and M. Avoli Network and Intrinsic Contributions to Carbachol-Induced Oscillations in the Rat Subiculum J Neurophysiol, September 1, 2001; 86(3): 1164 - 1178. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. van der Linden, F. Panzica, and M. de Curtis Carbachol Induces Fast Oscillations in the Medial but not in the Lateral Entorhinal Cortex of the Isolated Guinea Pig Brain J Neurophysiol, November 1, 1999; 82(5): 2441 - 2450. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. H Buhl, G. Tamas, and A. Fisahn Cholinergic activation and tonic excitation induce persistent gamma oscillations in mouse somatosensory cortex in vitro J. Physiol., November 15, 1998; 513(1): 117 - 126. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. L. McMahon, J. H. Williams, and J. A. Kauer Functionally Distinct Groups of Interneurons Identified During Rhythmic Carbachol Oscillations in Hippocampus In Vitro J. Neurosci., August 1, 1998; 18(15): 5640 - 5651. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. H. Williams and J. A. Kauer Properties of Carbachol-Induced Oscillatory Activity in Rat Hippocampus J Neurophysiol, November 1, 1997; 78(5): 2631 - 2640. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
