| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
The Journal of Neurophysiology Vol. 78 No. 5 November 1997, pp. 2742-2753
Copyright ©1997 by the American Physiological Society
Laboratoire de Neurophysiologie, Faculté de Médecine, Université Laval, Quebec G1K 7P4, Canada
 |
ABSTRACT |
|---|
 Abstract
 Introduction
Methods
Results
Discussion
References
|
|---|
Contreras, Diego, Niklaus Dürmüller, and Mircea Steriade. Absence of a prevalent laminar distribution of IPSPs in association cortical neurons of cat. J. Neurophysiol. 78: 2742-2753, 1997. The depth distribution of inhibitory postsynaptic potentials (IPSPs) was studied in cat suprasylvian (association) cortex in vivo. Single and dual simultaneous intracellular recordings from cortical neurons were performed in the anterior part of suprasylvian gyrus (area 5). Synaptic responses were obtained by stimulating the suprasylvian cortex, 2-3 mm anterior to the recording site, as well as the thalamic lateral posterior (LP) nucleus. Neurons were recorded from layers 2 to 6 and were classified as regular spiking (RS, n = 132), intrinsically bursting (IB, n = 24), and fast spiking (FS, n = 4). Most IB cells were located in deep layers (below 0.7 mm, n = 19), but we also found some IB cells more superficially (between 0.2 and 0.5 mm, n = 5). Deeply lying corticothalamic neurons were identified by their antidromic invasion on thalamic stimulation. Neurons responded with a combination of excitatory postsynaptic potentials (EPSPs) and IPSPs to both cortical and thalamic stimulation. No consistent relation was found between cell type or cell depth and the amplitude or duration of the IPSPs. In response to thalamic stimulation, RS cells had IPSPs of 7.9 ± 0.9 (SE) mV amplitude and 88.9 ± 6.4 ms duration. In IB cells, IPSPs elicited by thalamic stimulation had 7.4 ± 1.3 mV amplitude and 84.7 ± 14.3 ms duration. The differences between the two (RS and IB) groups were not statistically significant. Compared with thalamically elicited inhibitory responses, cortical stimulation evoked IPSPs with higher amplitude (12.3 ± 1.7 mV) and longer duration (117 ± 17.3 ms) at all depths. Both cortically and thalamically evoked IPSPs were predominantly monophasic. Injections of Cl
fully reversed thalamically as well as cortically evoked IPSPs and revealed additional late synaptic components in response to cortical stimulation. These data show that the amount of feed forward and feedback inhibition to cat's cortical association cells is not orderly distributed to distinct layers. Thus local cortical microcircuitry goes beyond the simplified structure determined by cortical layers.
Studies of the actions induced by the major excitatory and inhibitory neurotransmitters in the cortex, namely glutamate and Experiments were carried out on adult cats (2.5-3.5 kg) of either sex, anesthetized with pentobarbital sodium (35 mg/kg). All wounds and pressure points were infiltrated with lidocaine. Animals were paralyzed with gallamine triethiodide and artificially ventilated. End-tidal CO2 (3.5-3.7%) and heart rate were continuously monitored. Body temperature was maintained at 37-39°C. The depth of the anesthesia was maintained by additional doses of the same anesthetic, to keep a constant picture of high-amplitude low-frequency waves in the electroencephalogram (EEG).
Database and cellular identification
Results are based on 160 cortical cells from the cortical suprasylvian area 5. Cells were classified as regular spiking (RS, n = 132), intrinsically bursting (IB, n = 24), or fast spiking (FS, n = 4) according to their responses to depolarizing current pulses, following criteria established in vitro (Connors et al. 1982
Distribution of inhibition according to depth
We used two different approaches to compare the inhibitory input among cells at different depths. In the first type of analyses, we compared cells at different depths recorded from successive tracks in the same area, whereas, in the second type of analyses, we compared cells located at the same depth and recorded simultaneously <2-3 mm apart in the anteroposterior axis.
Intrinsically bursting cells
In several descents through area 5, IB cells were mostly encountered deeper than 0.7 mm (n = 19), but a smaller number of IB cells (n = 5) was located between 0.2 and 0.5 mm depth. The responses to thalamic stimulation of two superficial IB cells are illustrated in Fig. 4. The superficial cell in Fig. 4A was located at a depth of 0.2 mm and discharged typical spike bursts during spontaneous spindle oscillations (Fig. 4A2), similarly to the spindle-related bursts in deeply lying (1-1.5 mm) IB cells of area 5 (see Fig. 7 in Steriade et al. 1993b
Dual impalements
To directly compare the amplitude of the inhibitory input to different cells, triggered by the same stimuli, we recorded pairs of cortical cells simultaneously and studied their responses to thalamic and cortical stimulation. In the example of Fig. 5, two cells were recorded simultaneously at a similar depth, ~1.1 mm. Cell 1 was 2 mm anterior to cell 2 and therefore closer to the cortical stimulating electrode (see scheme in Fig. 1). Both cells responded to cortical and thalamic stimulation under different backgrounds of current injection, as indicated by the pulse protocols in Fig. 5. And both were corticothalamic cells because they were antidromically invaded from the LP nucleus with latencies of 0.6 ms for cell 1 and 0.5 ms for cell 2 (see insets at bottom). Antidromic invasion after thalamic stimulation was followed by an EPSP at 2.5 ms latency in cell 1, and 2.8 ms latency in cell 2 (Fig. 5, bottom panel, expanded detail indicated by arrow) that was crowned by action potentials and followed by a prominent IPSP in both cells. Cortical stimulation (top panels, Cx stim) triggered in both cells a longer lasting IPSP with a higher decrease in Rin, as compared with thalamic stimulation (bottom panels, Thal. stim). Noteworthy, cell 2 generated bursts of two to four action potentials at 230 Hz on depolarizing pulses applied at hyperpolarizing Vms (Fig. 5, Cx stim, cell 2, detail at right) and riding on a humplike potential, whereas smaller pulses elicited only passive responses.
Effect of stimulation intensity
Varying the intensity of stimulation revealed further differences in the local inhibitory circuit activated by cortical or thalamic volleys. By increasing the stimulation strength, the inhibitory input increased in all investigated cells (n = 12). Figure 8 depicts a pair of cells that were recorded simultaneously at depths of 1.2 mm (cell 1) and 1 mm (cell 2). Both cells had prominent and apparently similar IPSPs in response to cortical and thalamic stimulation. However, by changing the intensity of stimulation in a stepwise manner, some differences were revealed between the two cells (right column in Fig. 8 depicts expanded details). At the lowest intensity, cortical stimulation triggered a clear-cut biphasic IPSP in cell 1 (10 mV, total duration of 50 ms) but only a brief arrest of firing, without an apparent IPSP, in cell 2. Increasing the intensity of cortical stimulation revealed more steps in IPSP amplitude for cell 2 than for cell 1. In the case of thalamic stimulation, the smallest intensity did not evoke an IPSP but reduced firing frequency in both cells; thereafter, four steps of increased intensity revealed four different degrees of inhibition in cell 1, but a ceiling effect was observed in cell 2 after the second step.
Effect of intracellular Cl The IPSPs recorded after both types of stimulation had complex shapes due to the activation of mono- and polysynaptic pathways. An attempt was made to dissociate the Cl
The three major findings of this study are as follows. 1) Cortically as well as thalamically evoked IPSPs in neurons recorded from cortical association area 5 showed local differences in amplitude and duration that did not correspond to their localization in various layers. 2) IPSPs elicited by thalamic or cortical stimulation were almost completely reversed after intracellular injection of Cl Some methodological issues
The rather short latency of thalamically evoked antidromic responses in cortical neurons reported in this study (0.3-0.6 ms) may be surprising because layer VI cells have a small size and, presumably, a slow conduction velocity. However, several data from the recent and older literature explain that such short latencies are compatible with the conduction velocities and morphological characteristics of corticothalamic axons. 1) Axons originating in cells of cat's suprasylvian area 5 (where our experiments have been conducted) give rise to collaterals distributing to the striatum as well as the LP thalamic nucleus; the latter are thick and end in clusters of large boutons within the LP nucleus (Paré and Smith 1996 Absence of prevalent laminar distribution
In motor cortex in vitro (vanBrederode and Spain 1995
INTRODUCTION
Abstract
Introduction
Methods
Results
Discussion
References
-aminobutyric acid (GABA) (Connors 1992
; McCormick et al. 1993; Nicoll et al. 1990), mainly focused on the basic mechanisms involved in receptor activation and postsynaptic actions at the single-cell level. To understand the functioning of cortical microcircuitry, it is essential to determine the amount of excitatory and inhibitory inputs from specific thalamic and cortical projections. Excitatory inputs to cortical cells originate from a number of cortical and subcortical regions, while inhibitory inputs mainly arise in local smooth stellate cells that make symmetrical synaptic contacts, positive for the GABA-synthesizing enzyme glutamic acid decarboxylase (GAD) (reviewed in DeFelipe and Fariñas 1992). Inhibition is essential for the organization of receptive fields in the somatosensory (Mountcastle 1957) and visual (Hubel and Wiesel 1959) cortices, and tonic background inhibition is essential in the nervous system because its suppression may readily lead to epilepsy. In vitro, cortically evoked inhibitory postsynaptic potentials (IPSPs) in cortical cells have been shown to be due to the activation of GABAA and GABAB receptors (Connors et al. 1988), whereas, in vivo, the amplitude of GABAB responses is relatively much smaller (Contreras et al. 1996). One approach to understand functional microcircuits in the cortex is to quantify inhibition and excitation in relation to different cortical layers. Studies of visual cortex in vivo (Douglas and Martin 1991) and motor cortex in vitro (Kang et al. 1994; vanBrederode and Spain 1995) assessed the laminar distribution of inhibitory inputs and proposed general rules for the distribution of these inputs. We have explored the laminar distribution of inhibitory responses in a nonprimary neocortical area of cat, the association suprasylvian area 5, in which cortical and thalamic inputs are more distributed (Avendaño et al. 1985, 1988; Herkenham 1980; Jones 1985). Our study shows that the strength of inhibition associated to cortical or thalamic inputs does not depend on the laminar position, nor the electrophysiological type of neurons.
METHODS
Abstract
Introduction
Methods
Results
Discussion
References
. The depth of the pipettes was read on the display of the micromanipulator. A high-impedance amplifier (band-pass of 0-5 kHz) with active bridge circuitry was used to record and inject current into the cells. The signals were recorded on an eight-channel tape with band-pass of 0-9 kHz and posteriorly digitized at 20 kHz for off-line computer analysis. The EEG was recorded focally, from the vicinity of intracellularly recorded neurons, by means of coaxial electrodes with exposed areas of 0.2 mm separated by 0.6 mm. For monopolar EEG recordings, the indifferent electrode was placed in the neck muscles.
View larger version (24K):
[in a new window]
FIG. 1.
Physiological identification of cortical cells and measurements of inhibitory postsynaptic potential (IPSP). A: cell responding antidromically to either cortical or thalamic stimuli. In this and all following figures, cortical stimuli are abbreviated Cx and marked by a triangle, whereas thalamic lateral posterior (LP) stimuli are abbreviated Th and marked by a filled circle. Left panel: Cx stimuli at depolarized Vm ( 
55 mV, under +1 nA DC) and hyperpolarized Vm (95 mV, under 0.7 nA DC). Antidromic spike was followed by an excitatory postsynaptic potential (EPSP) and by an IPSP under depolarizing current. Hyperpolarization completely reversed the IPSP and blocked antidromic invasion. Right panel: responses to Th stimuli at the 2 Vm levels. The Cx- and Th-evoked antidromic and orthodromic spikes are expanded in insets. Two superimposed traces in each case. Below, the scheme shows the experimental paradigm used throughout the experiments. Also in the scheme are indicated the marginal gyrus (marg), the cruciate sulcus (s.cru), and the ectosylvian gyrus (ecto). Anterior is to the left, posterior is to the right. The 2 micropipettes were 2-5 mm apart in the anteroposterior axis of anterior suprasylvian gyrus (area 5), bipolar stimulating electrodes were placed 2-3 mm in front of the most anterior cell as well as in the thalamic LP nucleus. B: a different cell responded with an IPSP to thalamic stimulation. IPSP duration was measured at its half amplitude (thick horizontal line) and IPSP amplitude was measured at its peak (vertical dotted line) and from rest Vm (horizontal dotted line). See other details in text.
RESULTS
Abstract
Introduction
Methods
Results
Discussion
References
; McCormick et al. 1985) and in vivo (Nuñez et al. 1993). The recordings that were used for the database and analysis had stable membrane potentials (Vm), more negative than 55 mV (but Fig. 3A), for at least 15 min, and overshooting action potentials. Cells were located in layers 2 to 6 as estimated by reading the microdrive position display and using the cytoarchitectonic study of Hassler and Muhs-Clement (1964). This estimation of neuronal depth had an error of <15%, compared with the location of intracellularly stained neurons in areas 5 and 7 (Contreras and Steriade 1995; Steriade et al. 1993a). In a representative sample of 33 cells (including RS and IB cells), the resting membrane potential was 70 ± 1.6 mV, the amplitude of action potentials was 78 ± 1.4 mV, and the input resistance was 25.5 ± 4.1 M (estimated by applying square hyperpolarizing current pulses at the resting Vm).
View larger version (31K):
[in a new window]
FIG. 3.
Superficial and deep cells in area 5 may display similar IPSPs in response to thalamic LP stimulation. Responses of 2 regular spiking (RS) cells at 0.3 mm (A) and 1.3 mm (B) depth to thalamic stimulation. Vm was displaced by DC, and 2 traces superimposed for each Vm. Left voltage-current (V-I) plots (in A and B) show the Vm at rest (+), at the peak of the IPSP (15 ms in A and 30 ms in B; 
) and at 100 ms after the stimulus (
) for the different currents. Values of Vm were fit by regression lines (rest, 
; IPSP peak, ···; IPSP at 100 ms, - - -). Right plots (in A and B) show amplitude of the IPSP with respect to Vm at the peak () and at 100 ms ().
55 mV, +1 nA DC), cortical stimuli elicited an antidromic spike (0.3-ms latency) followed by an EPSP (1.2-ms latency) that was suprathreshold for firing an action potential (Fig. 1A, Cx, stimulus marked by a triangle). The EPSP was cut short by an IPSP that had a peak amplitude (at 18 ms) of 17.5 mV (measured from 60 mV) and a duration of 95 ms (at half amplitude). In the same neuron (Fig. 1A), cortical stimulation failed to display antidromic invasion at a hyperpolarized Vm (95 mV, 0.7 nA DC) that increased the amplitude of the EPSP and reversed the IPSP. The composite but constant nature of Cx-evoked PSPs was shown by the stereotyped shape of the superimposed traces in Fig. 1A at both Vms. However, some aspects of IPSP's shape may be ascribed to the polysynaptic nature of the activation and not necessarily to different GABAergic components. Thalamic stimulation (Fig. 1A, Th, stimulus marked by a filled circle) elicited in the same neuron a similar series of events, with antidromic invasion (0.3-ms latency) followed by orthodromic activation (1.7-ms latency). The Th-evoked IPSP had a peak amplitude of 13 mV (at 23 ms, measured from 60 mV) and a duration of 90 ms (at half amplitude). An example of an IPSP triggered by thalamic stimulation (Th, dot) is shown in Fig. 1B to illustrate how measurements were taken from the evoked IPSP: the dotted vertical line indicates the amplitude at the peak, selected by visual inspection, whereas the duration of the IPSP was taken at half amplitude (thick horizontal line). The IPSP depicted in Fig. 1B had a latency of 3.3 ms, the peak was at 16.6 ms, the duration was 67 ms, and it was not preceded by an EPSP. Maximum inhibitory responses were obtained by increasing the intensity of stimulation until reaching a value beyond which increasing intensity did not elicit a further increase in the amplitude of the response.
View this table:
TABLE 1.
Values of IPSPs evoked by cortical and thalamic stimulation in a sample of 60 cells from layers 2 to 6 in suprasylvian area 5
View larger version (25K):
[in a new window]
FIG. 2.
Distribution of thalamically evoked IPSPs' amplitude and duration in the depth of the cortex. Example of tracks from 5 different cats; the legend indicates the cat number (3, 11, etc.) for each. The most superficial cell was recorded at ~200 µm and the deepest at ~1,550 µm. Plots show the lack of a preferential distribution of IPSPs' amplitudes (top plot) or duration (bottom plot) in relation to the depth in the association cortical area 5. Absence of IPSPs around 1.1 mm is only apparent as IPSPs at that depth were found in several cats but not in those experiments selected for this figure.
65 mV) and durations (66 ms in the superficial cell, and 64 ms in the deeply lying cell). The degree of involvement of GABAA-activated conductances was inferred indirectly by the decrease in input resistance (Rin) during the peak of the IPSP, as compared with the Rin preceding the stimulation. The Rin was estimated from the slope of the regression line fitted to the voltage-current (V-I) plots (Fig. 3, left plots in A and B). Voltage values for the V-I plots were obtained from 10 ms before the stimulus (+), at the peak of IPSP (), and 100 ms after the stimulus (). At rest, the superficial cell (Fig. 3A) had a Rin of 21 M, which decreased to 5 M (72% drop) during the peak of the IPSP, and recovered to 20.5 M at 100 ms. In the deep cell (Fig. 3B), the Rin at rest was 26 M, at the peak of the IPSP it decreased to 9 M (65% drop), and was 20.5 M at 100 ms. By displacing the Vm with DC, the peak of the IPSP reversed at 66 mV in the superficial cell (Fig. 3A, right plot, ), and it reversed at 81 mV in the deep cell (Fig. 3B, right plot, ). The late phase of the IPSP (at 100 ms, ) reversed at 73 mV in the superficial cell and at 87 mV in the deep cell.
). Thalamic stimulation (Fig. 4A1) triggered an EPSP, with a latency of 3 ms, giving rise to an orthodromic spike that was followed by an IPSP with an amplitude of 6 mV (at a Vm of 60 mV) and a duration of 68 ms at half amplitude. Interestingly, spontaneous activity related to spindling generated strong bursting (Fig. 4A2), whereas thalamic electrical stimulation gave rise to single spikes followed by inhibition. This difference is possibly due to the fact that, in contrast to spontaneous spindle-related inputs, electrical stimuli to the thalamus readily recruit local cortical inhibition by backfiring corticothalamic axons. That cortical stimulation entrains stronger inhibition than thalamic stimulation is shown by IPSPs with higher amplitudes and longer durations triggered by cortical volleys (see Table 1). The other superficial IB cell fired spike bursts in response to depolarizing current pulses of increasing amplitude (Fig. 4B2). This cell was located at a depth of 0.25 mm and responded to thalamic stimulation with an EPSP at a latency of 4 ms, which was cut short by an IPSP of 8-mV amplitude (at 60 mV) and lasting for 70 ms. The postinhibitory burst in Fig. 4B may well originate from a low-threshold Ca2+ spike (LTS) but, at least in guinea pig frontal cortex, such LTSs were only found in deep layers and were absent from superficial layers (de la Peña and Geijo-Barrientos 1996). In addition to intrinsic factors, such as LTSs, synchronous excitatory inputs at the return of evoked hyperpolarizations may be taken in consideration, as was previously shown for the spontaneously occurring long-lasting hyperpolarizations of the slow oscillation (Contreras and Steriade 1995; Steriade et al. 1993a).
View larger version (22K):
[in a new window]
FIG. 4.
Intrinsically bursting cells were also located in superficial layers. A1: superimposed responses to thalamic stimulation of an intrinsically bursting (IB) cell at rest ( 64 mV, indicated by arrow) located at 200-µm depth, detail of the early response shown above. A2: spontaneous bursts during spontaneous spindle oscillations (displaced horizontally and vertically for clarity). B1: responses to thalamic stimulation of an IB cell located at 250-µm depth at different DC levels (Vm rest = 70 mV, indicated by arrow). B2: responses of the same cell to depolarizing pulses of increasing amplitude (from bottom to top).
View larger version (22K):
[in a new window]
FIG. 7.
Two deeply lying cells recorded simultaneously are characterized by different amounts of inhibitory input. Cell 1 (left column) and cell 2 (right column) were at depths of 1.2 and 1 mm, respectively. Cell 1 showed a smaller amplitude and duration IPSP to both cortical and thalamic stimulation. Stimuli applied during 2 different current levels, as indicated.
View larger version (42K):
[in a new window]
FIG. 5.
Simultaneously recorded neurons at similar depths may display dissimilar IPSPs. Cell 1 (left column) and cell 2 (right column) were 2 mm apart and were recorded simultaneously at a depth of 1.1 mm in area 5. Thalamic (bottom panel) and cortical (top panel) stimulation was applied on a background of different current injections (values are indicated). Both cells were antidromically activated from the thalamus and orthodromically from an adjacent focus in area 5 (expanded details indicated by arrow). Inhibitory responses in cell 1 had smaller amplitude and duration than in cell 2. Further details in text. In this and the following figures, small deflections that are concomitant with action potentials in the simultaneously recorded neuron are due to capacitive coupling.
), and an additional value was taken at 150-200 ms after stimulation (). Vm values were plotted against current, and Rin was read from the slope of regression lines fitted to the data. In cell 1, Rin decreased from 38 M at rest to 11 M (70% drop) at the peak of the cortically elicited IPSP, and recovered to 24 M at 200 ms after stimulation. The same cortical stimulus led to a decrease in Rin in cell 2, from 19 to 2 M (90% drop) at 30 ms, but the Rin was still only 12 M at 200 ms. Thalamic stimulation led to a drop in Rin of cell 1 from 44.1 to 14 M (68% drop) at the peak of the IPSP, and Rin was back to 37 M at 150 ms. In cell 2, thalamic stimulation brought Rin from 22.8 down to 6 M (71% drop) at the peak, but the Rin was still decreased (15 M) at 200 ms.
View larger version (23K):
[in a new window]
FIG. 6.
Voltage-current V-I plots from simultaneously recorded cells depicted in Fig. 5. Values of Vm were obtained from 10 to 30 ms before stimulation (+), 30 ms after stimulation, coincident with the peak of the IPSP ( ), and 150-200 ms after stimulation (). Vm values were plotted against DC current, and Rin was read from the slope of regression lines fitted to the data.
80 mV) and thalamic (78 mV) stimulation.
View larger version (61K):
[in a new window]
FIG. 8.
Varying the intensity of stimulation reveals differences in IPSPs between 2 cells recorded simultaneously at 1.2 mm (cell 1) and 1 mm (cell 2). Cells responded to cortical (top panel, Cx stim.) and thalamic (bottom panel, Th. stim.) stimulation with progressively increasing amplitudes of IPSPs. However, some differences among these neurons are seen with stepwise changes in stimulation strength to either cortex or thalamus (see text; details of the evoked IPSPs are expanded on the right column).
infusion
-dependent GABAA component from the K+-dependentGABAB component (n = 13), as described in vitro (Connors et al. 1988).
reversed the evoked IPSPs and revealed additional differences between cortical and thalamic stimulation. The cell depicted in Fig. 9 was located at 1.2 mm depth and was recorded with a 3-M KCl-filled pipette. Cortical stimulation applied after several minutes of impalement gave rise to an antidromic spike at 0.6-ms latency (see inset) followed by a strong depolarizing response at 4-ms latency, which triggered burst firing with spike inactivation. Bursting was not intrinsic but due to synaptic input, as indicated by the one-to-one correspondence between spikes and the individual components of the input, revealed by the hyperpolarizing pulse (detail in top panel). The initial response to cortical stimulation was followed, 80 ms later, by a secondary and similarly strong depolarizing response that is normally hidden by the IPSP. Thalamic stimulation (Fig. 9, bottom panel) gave rise to a single depolarizing event at a latency of 2.5 ms. Hyperpolarization blocked the occurrence of the spike and increased the amplitude of the thalamically evoked depolarizing synaptic response. To sum up, cortical stimulation set into action polysynaptic chains, whereas thalamic stimulation seemed to activate more circumscribed cortical zones.
View larger version (19K):
[in a new window]
FIG. 9.
Intracellular injection of Cl fully reversed the IPSPs. A cell was recorded at 1.2-mm depth, with a 3-M KCl-filled pipette. The cell was antidromically activated by cortical stimulation (Cx, top panel; see inset above) and orthodromically by thalamic stimulation (Th, bottom panel). No signs of IPSP remained after Cl leakage into the cell.
60 mV) and was reversed by hyperpolarizing pulses at 70 mV. Rin was evaluated from the slope of V-I plots (not shown) and dropped from 18.5 M at rest to 6 M during the peak of the IPSP, and was 13 M at 150 ms after the stimulus. Cell 2 was recorded simultaneously at a depth of 1.2 mm, with a KCl-filled pipette, and responded to the same thalamic stimuli, after several minutes of impalement, with an antidromic spike (0.5-ms latency, see detail in Fig. 10, bottom) followed by a high-amplitude and long-duration (70 ms) depolarizing potential (see detail in Fig. 10, top) that triggered burst firing. Thalamic stimulation applied during hyperpolarizing pulses failed to invade antidromically the cell and triggered a depolarizing synaptic response of increased amplitude.
View larger version (25K):
[in a new window]
FIG. 10.
Dual intracellular recordings with KCl- and KAc-filled pipettes. Cell 1 was recorded with KAc pipette and showed a prominent IPSP in response to thalamic stimulation (see inset). Simultaneously recorded cell 2 showed no IPSP on the same stimuli after Cl leaked into the cell (see inset). Cell 2 was antidromically activated by thalamic stimulation, at 0.5-ms latency (detail in bottom panel).
injections. This is shown for a cell located at a depth of 0.8 mm, recorded with a 1.5-M KCl-filled pipette (Fig. 11). The top and bottom panels in Fig. 11 show two depolarizing current pulse of different amplitudes (values are indicated), during which thalamic stimuli were delivered. Superimposed traces show the evoked IPSP immediately after impalement and ~15 min after impalement. Before the IPSP reversal by the leak of Cl inside the cell, the IPSP had and amplitude of 4.2 mV (peak at 30 ms) and 130 ms in duration, measured from 60 mV. Its reversal was at 70 mV. The Rin of the cell was 26 M, dropped to 12 M at the peak of the IPSP, and was 16 M at 120 ms. After several minutes of recording, hyperpolarizing IPSPs were no longer evoked, and stimulation triggered a burst of action potentials. The burst was followed by a pause, whose duration depended on the amount of current being applied (40 ms in top panel, with a pulse of +0.5 nA, and 160 ms in the bottom panel, with a +0.2-nA pulse). Thereafter the firing resumed due to the application of depolarizing current.
View larger version (38K):
[in a new window]
FIG. 11.
Progressive reversal of thalamically evoked IPSPs with Cl . Cell recorded with a1.5-M KCl-filled pipette showed a progressive reversal of the IPSPs. Top and bottom panels show, for different background current levels (depolarizing pulses of +0.5 and +0.2 nA), superimposed responses to thalamic stimulation applied immediately after impalement and a few minutes after impalement.
DISCUSSION
Abstract
Introduction
Methods
Results
Discussion
References
. 3) IB cells from all layers showed robust inhibitory responses, not different from those recorded in RS cells.
). It is plausible that the thick corticothalamic axons described in this study conduct at high velocities and explain the short latencies found in our paper. 2) Another, nonexclusive possibility is that our stimulus spread to the centrum medianum (CM) nucleus that is located at no more than 1-1.5 mm medial to LP (or the stimulus activated cortico-CM passing axons). In a previous study, it was found that the cortical cells in areas 5 and 7, projecting to CM, have high conduction velocities, up to 33 m/s, much faster than in other corticothalamic systems (Steriade et al. 1978).
; Jahnsen and Llinás 1984).
et al. 1966; Renaud et al. 1974). Besides, this was not a major concern, because the objective of the present study was not to differentiate between corticocortical and thalamocortical pathways, but to determine whether there is a distinct laminar distribution of inhibitory inputs. Two differences were evident, however, between cortically and thalamically evoked IPSPs: 1) cortically elicited IPSPs were of higher amplitude, longer duration, and associated to larger increases in conductance than those evoked by thalamic stimulation; and 2) only thalamic stimulation was capable of eliciting IPSPs in isolation.
injections (see Figs. 10 and 11), it is likely that intrinsic currents may participate in shaping the synaptic responses. No attempt was made to dissociate the intrinsic components in the synaptic responses investigated here.
), strong inhibitory responses to local stimulation consisted of GABAA-B components, were present in layers 2 and 3, and were weak or absent in layer 5. This difference was attributed, at least in part, to the strong Ih in layer V neurons. In visual cortex in vivo (Douglas and Martin 1991), both types of GABAergic responses were present in all layers, butGABAA IPSPs were more pronounced in deep layers, whereas IPSPs in superficial layers developed more slowly and lasted longer. These data led to the proposal of a "canonical" circuit for the visual cortex, despite the rich interconnectivity between layers (Gilbert and Wiesel 1979; Kisvárday et al. 1987; Martin and Whitteridge 1984; Somogyi et al. 1983). Our data suggest that the microcircuitry in the association cortex is more complicated and that the amount of inhibition associated to thalamic or cortical inputs is determined by the local microcircuitry in a manner that remained beyond the methods of exploration used here.
). In our study IPSPs were never evoked alone on cortical stimulation, but they could be triggered in isolation after thalamic stimulation (see Figs. 1B and 3). This result may suggest a lower threshold for inhibition than for excitation in thalamocortical pathways, compared with corticocortical ones. However, at high-intensity stimulation, cortically elicited IPSPs were consistently of higher amplitude and longer duration and were accompanied by larger increments in membrane conductance (see dual intracellular recordings in Figs. 5 and 7). In those cases in which IPSPs were observed in isolation, the latency was longer than 3 ms, which is compatible with at least a bisynaptic pathway, therefore involving intermediary intracortical steps.
) and in vitro (Gil and Amitai 1996). As well, the IPSPs were reversed almost completely by Cl. This result is at variance with the expected biphasic (GABAA- andGABAB-mediated) IPSP that has been described in vitro (Connors et al. 1988; vanBrederode and Spain 1995). One possible explanation that has recently been advanced (Contreras et al. 1996) is that the evoked GABAA IPSP shuts off cellular activity in a large proportion of cells, thus leading to a generalized phenomenon of disfacilitation during which K+ currents, responsible for the resting Vm, dominate the membrane behavior. It was then hypothesized that the apparent absence of GABAB components is due to the overwhelming effect of disfacilitation. Such a scenario may only be observable in a situation, such as in vivo, with important spontaneous activity. In the experiments presented here, under barbiturate anesthesia, background activity is smaller than under ketamine-xylazine, and the additional explanation may be offered that GABAB IPSPs are masked by the decreased Rin of a long-lasting GABAA IPSP. It has been shown in vitro (Thompson and Gäwhiler 1992) that pentobarbital sodium increases the decay time constant of GABAA-mediated currents, consequently prolonging GABAA-mediated IPSPs, as well as that long-lasting GABAA IPSPs may be due to polysynaptic activity. Long-lasting (0.1-0.4 s) monophasic IPSPs, with peak amplitudes at ~30 ms, have been described in vivo (Pollen and Lux 1966; Renaud et al. 1974); in those studies the percentage decrease in Rin at the peak of the IPSP varied considerably, from <50 to 100% (see also Krnjevi et al. 1966) (up to 60%). This variation, including our own results (60-90% according to the stimulated pathway), probably reflects the number and location of synapses activated, which in turn depends on the intensity and location of the stimulation. The time course and amplitude of the IPSPs presented in the present study is in full agreement with those earlier studies.
), some basic questions are as follows. 1) Is there a laminar arrangement of inputs that predicts a laminar organization of excitatory and/or inhibitory responses? 2) What is the distribution of symmetric and asymmetric synapses over the membrane of cells located at different depths? And 3) do thalamic or cortical inputs contact preferentially a specific cell type? A classical description of laminar distribution of thalamic inputs (Lorente de Nó 1938) separated a "specific" system with terminal arborizations densely packed in layers 3 and 4 and an "unspecific" system terminating sparsely in layers 1 and 6. Thalamic inputs to primary sensory areas, such as the visual cortex, show characteristic laminar differences in the distribution of EPSPs and IPSPs (Ferster and Lindström 1983). The LP-pulvinar nuclear complex projects to layers 4 and 1 of area 5 in the cat (Robertson 1978). Anatomic data indicate that there is a widely distributed layer 1 projecting system in the thalamus, interspersed among neurons projecting to deeper layers (Avendaño et al. 1990; Steriade and Deschênes 1984). Despite the diversity of inhibitory inputs to cortical cells, the large decreases in Rin associated with the IPSPs indicate that a large proportion of the inhibitory action occurs at or near the cell soma. Therefore neither anatomic nor the present physiological data indicate any preferential distribution of inhibitory inputs to distinct layers in the association cortex. In contrast to successful studies in primary cortices (mainly motor and visual), association areas defy more global approaches and require for its understanding combined physiological and anatomic studies at the single-cell level.
| |
ACKNOWLEDGEMENTS |
|---|
This work was supported by grants from the Medical Research Council and Human Frontier Science Program. D. Contreras was a Ph.D. student supported by the Savoy Foundation. N. Dürmüller was a postdoctoral fellow.
| |
FOOTNOTES |
|---|
Address for reprint requests: M. Steriade, Laboratoire de Neurophysiologie, Faculté de Medecine, Université Laval, Quebec G1K 7P4, Canada.
Received 23 May 1997; accepted in final form 21 July 1997.
| |
REFERENCES |
|---|
|
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
, K.,
RANDIC, M.,
STRAUGHAN, D.
Nature of a cortical inhibitory process.
J. Physiol. (Lond.)
184: 49-77, 1966.This article has been cited by other articles:
|
|
 |
|
  M. Rudolph, J. G. Pelletier, D. Pare, and A. Destexhe Characterization of Synaptic Conductances and Integrative Properties During Electrically Induced EEG-Activated States in Neocortical Neurons In Vivo J Neurophysiol, October 1, 2005; 94(4): 2805 - 2821. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Butovas and C. Schwarz Spatiotemporal Effects of Microstimulation in Rat Neocortex: A Parametric Study Using Multielectrode Recordings J Neurophysiol, November 1, 2003; 90(5): 3024 - 3039. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. S. Jones, K. D. MacDonald, B. Choi, F. E. Dudek, and D. S. Barth Intracellular Correlates of Fast (>200 Hz) Electrical Oscillations in Rat Somatosensory Cortex J Neurophysiol, September 1, 2000; 84(3): 1505 - 1518. [Abstract] [Full Text] [PDF]
|
|
|
|
