| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Information Physiology, National Institute for Physiological Sciences, Okazaki, Japan
Submitted 20 March 2006; accepted in final form 13 July 2006
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
). Thalamocortical (TC) cells in the ventrobasal thalamus project their axons to layer 4 of the barrel cortex (Jensen and Killackey 1987; Killackey 1973). Functionally, the TC cells are connected to excitatory regular-spiking (RS) cells and inhibitory fast-spiking (FS) cells in cortical layer 4 (Bruno and Simons 2002; Simons 1978; Swadlow 1995).
Feedforward inhibition is a salient feature observed in the somatosensory thalamocortical circuit (reviewed in Swadlow 2003). Using thalamocortically connected slices (Agmon and Connors 1991; Fig. 1A), synaptic organization governing the feedforward inhibition has been studied in vitro, at the level of synaptic potentials and currents. Thalamic stimulation produces feedforward inhibitory postsynaptic potentials (IPSPs) in layer 4 RS cells (Agmon and Connors 1992). Minimal stimulation of presumed single TC fibers generates all-or-none excitatory postsynaptic potentials (EPSPs) in both RS and FS cells in layer 4 (Beierlein et al. 2003; Gibson et al. 1999; Gil et al. 1999). Action potentials in layer 4 FS cells generate IPSPs in layer 4 RS cells (Beierlein et al. 2003; Sun et al. 2006).
|
; Lin et al. 1985) and not all GABAergic interneurons are FS cells (Amitai et al. 2002; Gibson et al. 1999). Thus excitatory RS cells are much larger in number than inhibitory FS cells. In this study, we therefore investigated the connection rules of the feedforward inhibitory circuit from multiple TC cells to multiple RS cells by FS cells, at the level of synaptic potentials.
A recent study detected divergent synaptic inputs of TC fibers onto RS and FS cells, in which paired patch-clamp recordings of RS and FS cells were made and the minimal stimulation was then applied to presumed single TC fibers (Gabernet et al. 2005). The authors observed tightly correlated occurrence of all-or-none EPSC responses in both RS and FS cells and concluded that the correlated events were caused by activation of the same TC fiber. Such a correlation-based experiment was also reported in the lateral geniculate nucleus (LGN) of the thalamus (Blitz and Regehr 2005).
Here we used the same correlation-based paradigm, to address the synaptic organization of the multiple-to-multiple thalamocortical feedforward circuit. Our quantitative analysis revealed that 1) nearly all of the TC fibers that generated excitatory inputs onto RS cells also generated excitatory inputs onto adjacent FS cells; 2) adjacent pairs of RS cells also received divergent excitatory inputs from the same TC fiber; and 3) FS cells received convergent excitatory inputs from multiple TC cells and sent divergent inhibitory outputs to multiple RS cells.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Experiments were performed using C57/BL6 mice (postnatal days 12–16). All experimental procedures were in accordance with the institutional guidelines concerning the care and handling of experimental animals and were approved by the Animal Research Committee of our institute. Animals were anesthetized by halothane (Takeda Chemical Industries, Osaka, Japan) and then killed by decapitation. The brain was rapidly removed and immediately placed in a chilled, artificial cerebrospinal fluid (ACSF). The ACSF contained (in mM): NaCl 124, KCl 2.5, NaH2PO4 1.1, MgSO4 1.2, CaCl2 2.5, NaHCO3 26, and glucose 12. It was oxygenated by a gas mixture containing 95% O2-5% CO2. Thalamocortical slices (300 µm thickness), where the thalamic ventrobasal region and the barrel cortex were synaptically connected, were prepared according to Agmon and Connors (1991). The slices were transferred into a holding tank, incubated at 32°C for 30 min, and then maintained at room temperature until just before recording.
Individual slices were transferred to a submerged recording chamber and were continuously perfused with the ACSF kept at 30–32°C. The slices were visualized using infrared differential interference contrast (IR-DIC) microscopy (Axioskop FS, Carl Zeiss, Jena, Germany) equipped with a Newvicon tube camera (C2400-07, Hamamatsu Photonics, Hamamatsu, Japan). Three adjacent cells in cortical layer 4 were randomly selected for electrophysiological recording. The intercellular distance, measured from center to center, was <40 µm.
Membrane potentials of three cortical cells were simultaneously recorded in current-clamp mode using three patch-clamp amplifiers, two Axoclamp-2B amplifiers (Molecular Devices, Sunnyvale, CA), and an EPC-7 amplifier (HEKA Elektronik, Lambrecht, Germany). Patch pipettes were filled with an intracellular solution containing (in mM): K-methanesulfonate 135, NaCl 6, EGTA 0.2, HEPES 10, Mg-ATP 4, Na3-GTP 0.3, and phosphocreatine-Tris 10 (pH 7.3, 290–295 mOsm). The pipette resistance was 4–6 M
with the intracellular solution. Liquid junction potentials were not corrected. All electrical signals were low-pass filtered at 3 kHz and then digitally sampled at 10 kHz using an ITC-16 interface (Instrutech, Port Washington, NY) that was controlled by IgorPro (Wavemetrics, Lake Oswego, OR).
Only cortical cells with resting membrane potentials more negative than –50 mV and overshooting action potentials were accepted for this study. Characterization of layer 4 cortical cells (RS cells and FS cells) was based on the firing patterns (Fig. 1B). The RS cells had a mean resting membrane potential of –69.0 ± 0.4 mV (n = 104) and a mean membrane input resistance of 432.9 ± 11.9 M (n = 104), whereas the FS cells had a mean resting membrane potential of –65.6 ± 0.5 mV (n = 52) and a mean membrane input resistance of 81.7 ± 3.4 M (n = 52). Thalamocortical EPSPs were measured at the holding potential of –73 mV, whereas FS-evoked IPSPs in RS cells were measured at the holding potential of –53 mV.
Minimal stimulation of thalamic fibers
A custom-made bipolar electrode was used for minimal stimulation of presumed single TC axons. The bipolar electrode was made by gluing two monopolar tungsten microelectrodes (FHC, Bowdoinham, ME). The distance between their tips was about 250 µm. For local electrical stimulation, the parallel bipolar electrode was placed perpendicular to the surface of the slices: one tip of the bipolar electrode was pushed into the slices, whereas the other tip was placed away from the surface of the slices.
The stimulating electrode was placed within the ventrobasal thalamus. Electrical stimulation was applied as a cathodal pulse with 100-µs duration through the electrode tip pushed into the slices. The electrical pulse was generated by a stimulus isolation unit (ISO-Flex, AMPI, Jerusalem, Israel). The stimulus intensity was adjusted in each experiment of minimal stimulation (range: 40–250 µA), to evoke all-or-none responses with about 50% success rates. At the threshold intensity, 20–60 trials of stimuli (0.1–0.3 Hz) were applied in each experiment of minimal stimulation. We accepted the following responses as minimal stimulation-evoked EPSPs: 1) when the responses were clearly all-or-none, 2) the responses had a constant EPSP latency, and 3) a small change of stimulus intensity changed the success rates of the EPSPs but did not change the EPSP amplitudes.
In the correlation-based paradigm (Fig. 2), "reference" connections were critical for us to judge whether single TC fibers were stimulated. Therefore our criterion to establish the reference connections was that minimal stimulation-evoked responses in cortical cells had EPSP amplitudes >2 mV. Such EPSP responses exhibited clearly distinguishable success and failure events, so that we could easily determine the successes and failures by eye. There were no limitations on EPSP amplitude in "test" connections. When more than one cortical cell could be regarded as reference cells, we regarded both RS and FS cells as reference cells and then individually analyzed the presence of divergent thalamocortical inputs (see Fig. 4, A1 and A2).
|
|
|
We quantitatively analyzed the correlation-based paradigm in the following way. Success and failure events in reference cells could be easily determined by eye as described above, whereas success and failure events in test cells were determined by the threshold obtained from 1.6 x RMS (root mean square) of the baseline noise in test cells (Beierlein et al. 2003; Markram et al. 1997). In this study, our criterion for correlated responses was that all of the success events in the reference cells were accompanied by the success events in the test cells (i.e., 100% match of the success events). The complete match indicates that we strictly detected the divergence of thalamocortical inputs.
To examine the firing patterns of the cortical cells, repetitive firing was elicited by a 500-ms positive current injection. The injected current was adjusted to evoke about 50 Hz of initial spike-frequency (i.e., the inverse of the interval between the first and second spikes). The spike-frequency adaptation during the repetitive firing was calculated as the ratio of the average of the last four spike frequencies to the initial spike frequency. The spike width of action potentials was calculated at the half-amplitude of the first spike during the repetitive firing. The amplitude of afterhyperpolarization (AHP) after action potentials was calculated as the minimal membrane potential between the first spike and the second spike during the repetitive firing, which was subtracted from the membrane potential of the first spike threshold.
The synaptic latency of the EPSP onset was determined as the time from stimulus artifact to 5% of the EPSP amplitude. Amplitudes of individual EPSP responses during the correlation-based paradigm were measured as the individual peak values (searched during 10 ms just after the EPSP onset), subtracted from the individual baseline values (mean membrane potential during 1 ms just before the EPSP onset). Thalamocortical EPSP responses were discarded if feedforward IPSPs or polysynaptic EPSPs were observed. All analyses were performed using IgorPro. All pooled data were represented as means ± SE except as noted.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
We identified excitatory RS cells and inhibitory FS cells in layer 4 of the somatosensory barrel cortex, based on the firing patterns (Beierlein et al. 2003; Gibson et al. 1999). Figure 1B shows an IR-DIC image of a triple patch-clamp recording of cortical cells (two RS cells and one FS cell) and their firing patterns. RS cells displayed strong spike-frequency adaptation during repetitive firing (adaptation ratio: 0.34 ± 0.01, n = 104), broad spike width of action potentials (1.16 ± 0.02 ms, n = 104), and small AHP amplitudes after action potentials (4.4 ± 0.3 mV, n = 104). To record FS cells, we targeted neurons with relatively large somata (Fig. 1B; Gibson et al. 1999; Simons and Woolsey 1984). FS cells displayed little spike-frequency adaptation (adaptation ratio: 0.90 ± 0.03, n = 52), narrower spike width (0.45 ± 0.01 ms, n = 52), and deeper AHP (15.9 ± 0.4 mV, n = 52).
A minimal stimulation paradigm has frequently been used to assess the unitary responses in somatosensory thalamocortical synapses (Beierlein et al. 2003; Gabernet et al. 2005; Gibson et al. 1999; Gil et al. 1999). In this paradigm, the extracellular stimulation intensity at the thalamus is minimized and adjusted to evoke clear all-or-none synaptic responses in cortical cells. The success events evoked by the threshold stimulation are presumed to originate from activation of a single presynaptic TC fiber. We recorded such all-or-none responses in RS or FS cells, using the minimal stimulation paradigm (Fig. 1C). The minimal stimulation-evoked EPSPs in RS cells had a mean amplitude of 3.7 ± 0.4 mV (n = 33), whereas those in FS cells had a larger amplitude (6.9 ± 0.7 mV, n = 31). The mean synaptic latencies from thalamic stimulation to the EPSP onset in cortical cells were 3 to 4 ms (3.6 ± 0.1 ms in RS cells, n = 33; 3.2 ± 0.1 ms in FS cells, n = 31), which are consistent with the previous report (Beierlein et al. 2003).
Detection of functional divergence of thalamic fibers onto multiple cortical cells
Gabernet et al. (2005) recently detected divergent synaptic inputs of TC fibers onto RS and FS cells. They made paired recordings of RS and FS cells and then applied the minimal stimulation to presumed single TC fibers. They reported that, in particular examples, the minimal stimulation elicited tightly correlated occurrence of all-or-none responses in both RS and FS cells, concluding that the RS and FS cells received EPSC inputs from the same TC fibers. In the present study, we quantitatively examined the probability of detecting the divergent synaptic inputs, using the same correlation-based paradigm.
We first made patch-clamp recordings from three adjacent cortical cells, located deeply from the cut surface of the brain slice. We then moved a stimulating electrode within the ventrobasal thalamus and searched for a spot where clear all-or-none EPSPs were observed in one of the three cortical cells. When we observed all-or-none EPSPs in a cell, we regarded it as the "reference" cell and the two other cells as the "test" cells. Interestingly, we frequently observed all-or-none responses in both the reference and test cells, even though we applied the minimal stimulation only to the reference cell ("All" in Fig. 2A). In addition to the identical threshold, the occurrence of the all-or-none EPSPs was tightly correlated ("Individual" in Fig. 2A and Fig. 2B). This indicates that the reference and test cells received EPSP inputs from the same TC fiber. Figure 2, B and C shows how to determine whether the EPSP responses in the reference and test cells are correlated.
Quantitative analysis of the divergent thalamocortical inputs
Using the correlation-based paradigm, we not only observed thalamocortical divergence onto RS and FS cells (Fig. 2), but also observed thalamocortical divergence onto pairs of RS cells that had not been previously reported (Fig. 3). We then quantitatively examined the connectivity of TC fibers onto RS and FS cells (Fig. 4A) and the connectivity of TC fibers onto pairs of RS cells (Fig. 4B). Because we simultaneously recorded three cortical cells (RS1, RS2, and FS cells) in each experiment, two divergent inputs onto RS and FS cells (RS1–FS pair and RS2–FS pair) and one divergent input onto pairs of RS cells (RS1–RS2 pair) could be analyzed from each experiment. Pairs of cortical cells, neither of which received thalamocortical EPSP inputs, were excluded from the quantitative analysis because we suspected that TC axons were severed during the slicing procedures.
The divergent inputs from TC fibers to RS and FS cells were first analyzed (Fig. 4A). When we regarded RS cells as the reference cells (Fig. 4A1), 96.6% (28/29) of the RS–FS pairs received EPSP inputs from the same TC fibers (filled circles in Fig. 4A1), whereas we found that in only 3.4% (1/29) of the RS–FS pairs, the reference RS cell alone received EPSP inputs (an open circle in Fig. 4A1). When we regarded FS cells as the reference cells (Fig. 4A2), 53.7% (29/54) of the RS–FS pairs received EPSP inputs from the same TC fibers (filled circles in Fig. 4A2), whereas 46.3% (25/54) of the RS–FS pairs received EPSP inputs onto only the reference FS cells (open circles in Fig. 4A2). These results indicate that nearly all of the TC fibers generated EPSP inputs onto both RS and FS cells (filled circles in Fig. 4, A1 and A2) or onto only FS cells (open circles in Fig. 4A2), but not onto only RS cells (an open circle in Fig. 4A1). These results also indicate that, when a TC fiber generated excitatory inputs onto RS cells, the TC fiber also generated divergent excitatory inputs onto adjacent FS cells with high fidelity (96.6%; Fig. 4A1).
The divergent inputs from TC fibers to adjacent pairs of RS cells were then examined (Fig. 4B). We found that adjacent RS pairs also exhibited tightly correlated EPSP responses using the correlation-based paradigm (Fig. 3). However, the probability of detecting the divergent inputs was not high. In 43.5% (10/23) of the RS–RS pairs, two RS cells received EPSP inputs from the same TC fibers (filled circles in Fig. 4B). In the rest of the RS–RS pairs (56.5%; 13/23), only one of the RS–RS pairs received EPSP inputs (open circles in Fig. 4B). These results indicate that adjacent RS cells received divergent excitatory inputs from the same TC fiber. However, these results also indicate that the probability of the RS–RS pairs receiving the divergent inputs (43.5%; Fig. 4B) was lower than the probability of the RS–FS pairs receiving the divergent inputs (96.6%; Fig. 4A1), when we regarded RS cells as the reference cells.
Convergent synaptic inputs from multiple thalamic fibers onto single FS cells
Figure 4A shows that, even if a TC fiber did not generate EPSP inputs onto RS cells, the TC fiber generated EPSP inputs onto FS cells. This implies a high probability of convergent excitatory inputs from multiple TC cells onto FS cells. To directly address this issue, we repeatedly used the correlation-based paradigm in the same slice (Fig. 5).
|
Using this method, we quantitatively examined the convergent thalamocortical inputs onto FS cells (Fig. 5C). We recorded from seven triplets (two RS cells and one FS cell), and stimulated at least two (two to three) TC fibers in each triplet. We confirmed the independent stimulation of different TC fibers, based on whether the correlation-based paradigm at different spots produced different patterns of EPSP amplitudes in RS1 and RS2 cells (see Fig. 5B2 and the legends for details). Six of seven FS cells we examined (85.7%) received EPSP inputs from all of the stimulated TC fibers (Fig. 5C). The remaining one FS cell received EPSP inputs from two of three stimulated TC fibers (no. 7 in Fig. 5C). These results indicate a high probability of convergent synaptic inputs from multiple TC fibers onto FS cells.
Divergent synaptic inputs from single FS cells to multiple RS cells
Inhibitory connections from FS cells to RS cells are vital parts of the feedforward inhibitory circuit from the thalamus to layer 4 of the barrel cortex. Previous reports using paired whole cell recordings revealed inhibitory connections from single FS cells to single RS cells (Beierlein et al. 2003; Gabernet et al. 2005; Sun et al. 2006). We finally examined the divergent synaptic inputs from single FS cells to adjacent pairs of RS cells quantitatively, using simultaneous whole cell recordings from three cells (Fig. 6).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Technical considerations
The correlation-based paradigm in this study was used to examine whether two postsynaptic cells received divergent synaptic inputs from the same presynaptic fiber. Using this paradigm, Gabernet et al. (2005) reported that RS and FS cells in the barrel cortex received divergent EPSC inputs from the same TC fiber. In the LGN of the thalamus, Blitz and Regehr (2005) examined whether direct EPSCs and feedforward IPSCs in a thalamic relay cell originated from the same retinogeniculate fiber. Thus the correlation-based paradigm has been recognized as a useful technique to detect divergent synaptic inputs.
In general, synaptic connectivity would be underestimated in slice preparations because some axons are severed during the slicing procedure. Long-range TC axons in thalamocortical slices (Agmon and Connors 1991) may be especially exposed to the axonal cutting. Despite this potential for underestimation, we observed an extremely high probability (96.6%) of divergent inputs from TC fibers onto RS and FS cells, when TC 
RS connections were established (Fig. 4A1). This may be attributable to our correlation-based paradigm in which we established the reference connections. We did not use slice preparations where we could not find any reference connections. The procedures enabled us to exclude the slices in which thalamocortical axons were severed, which consequently minimized the underestimation.
We also detected divergent synaptic inputs from TC fibers onto pairs of RS cells (Figs. 3 and 4B). However, the connectivity was not high (43.5%). Although it remains unclear whether the connectivity onto pairs of RS cells is underestimated or biologically low, we suspect a possibility of the underestimation based on the following two reasons. First, thalamocortical synapses are distributed on both proximal and distal dendrites of spiny stellate cells (a major type of RS cell), but on the somata of inhibitory interneurons (Benshalom and White 1986; Keller and White 1987; White et al. 1984). Distally located synapses onto RS cells might be subject to the axonal cutting. Second, the correlation-based paradigm in the present study could not detect small (roughly <1 mV) EPSPs as observed in RS cells (e.g., "RS2" in Fig. 2, A and C), which also would result in an underestimation of the connectivity.
The probability of detecting EPSP inputs onto RS cells was 53.7% when FS cells were the reference cells (Fig. 4A2), but was lower when RS cells were the reference cells (43.5%; Fig. 4B). As discussed above, we suspect that the different probabilities were also derived from our correlation-based paradigm, which could not detect small EPSP amplitudes. In the case of FS cells as reference cells (Fig. 4A2), EPSP amplitudes in only the test RS cells have to be large enough to be judged as correlated responses (i.e., roughly >1 mV). On the other hand, in the case of RS cells as reference cells (Fig. 4B), EPSP amplitudes in both of the reference and test RS cells must be sufficiently large.
In this study, we placed an extracellular stimulating electrode in the thalamus, to activate TC fibers. Extracellular thalamic stimulation not only activates TC fibers orthodromically, but also activates corticothalamic (CT) fibers antidromically (reviewed in Castro-Alamancos 2004). Thus there is a possibility that some of thalamus-evoked EPSPs in our recordings are derived from the activation of CT fibers. However, Beierlein et al. (2003) reported that the TC fiber–evoked EPSPs had synaptic latencies of about 3 ms, whereas the CT fiber–evoked EPSPs had longer synaptic latencies of 6 to 7 ms. The synaptic latencies of the EPSPs we recorded (3 to 4 ms) were consistent with those of the TC fiber–evoked EPSPs, suggesting that we mainly stimulated TC fibers.
The stimulus intensity of the minimal stimulation in this study (range: 40–250 µA) was relatively higher than that in previous reports (range: 5–100 µA; see Beierlein et al. 2003; Gabernet et al. 2005). Nevertheless, limited numbers of TC fibers appeared to be locally stimulated in our recordings for the following reasons. When the minimal stimulation was applied to one of three simultaneously recorded cells (see "RS1 Reference" in Fig. 2A), we frequently observed correlated EPSP responses in the other two cells (see "RS2" and "FS" in Fig. 2A). However, we could not observe any additional synaptic responses in the other two cells. If many TC fibers were stimulated, we must have observed such additional synaptic responses in the other two cells.
Synaptic organization of the thalamocortical connection
In this study, we frequently observed that adjacent cortical cells received divergent inputs from the same TC fiber (Figs. 2–4). These divergent thalamocortical inputs, based on our electrophysiological study, are consistent with previous morphological studies. Single thalamic axons are extensively ramified within single barrels (extent of the axon arbors, about 200–500 µm in rats; Jensen and Killackey 1987). The dendritic field span of rat layer 4 neurons (spiny stellate cells and FS cells) is about 200 µm (Amitai et al. 2002; Lübke et al. 2000; Simons and Woolsey 1984). Thus the extensive ramification of single TC axons is consistent with our finding that adjacent cortical cells frequently received divergent inputs from the same TC fiber.
Properties of neuronal firing have been used for classification of cortical inhibitory interneurons (Kawaguchi 1995). In layer 4 of the rodent barrel cortex, nonadapting inhibitory interneurons are classified as FS cells, whereas two types of adapting inhibitory interneurons have been reported: low-threshold spiking (LTS) cells (Beierlein et al. 2003; Deans et al. 2001; Gibson et al. 1999) and regular-spiking nonpyramidal (RSNP) cells (Beierlein et al. 2002; Porter et al. 2001; Sun et al. 2006). However, these previous studies also suggested that FS cells are major mediators of thalamocortical feedforward inhibition. Thalamocortical EPSPs in LTS cells are much smaller in amplitude than those in FS cells (Beierlein et al. 2003; Gibson et al. 1999). Although thalamocortical EPSPs in RSNP cells have amplitudes comparable to those in FS cells (Porter et al. 2001), unitary IPSPs from RSNP cells to spiny cells have small amplitudes (Sun et al. 2006).
Connectivity between thalamic and cortical cells has been well studied using in vivo extracellular recordings, in which thalamic and cortical spikes were simultaneously recorded and then the connectivity was assessed by cross-correlation analysis of the spikes. The cross-correlation analysis revealed that connectivity between TC and FS cells was significantly higher than that between TC and RS cells (Bruno and Simons 2002) and that FS cells received highly convergent inputs from multiple TC cells (Swadlow and Gusev 2002). However, thalamocortical convergence has not been studied at the finer level of synaptic potentials. Our present study using in vitro intracellular recordings (Fig. 5) was consistent with previous studies using in vivo extracellular recordings.
In this study, we examined divergent thalamocortical inputs onto nearby (about 40 µm) cortical cells, but did not examine those onto distant cortical cells throughout single barrels (about 150–300 µm in diameter in mice; Woolsey and van der Loos 1970). Thus our studies provide information about local- and fine-scale thalamocortical connectivity, rather than large-scale connectivity (e.g., studies of the size of whisker receptive fields in vivo: Armstrong-James and Fox 1987; Moore and Nelson 1998; Zhu and Connors 1999). Our approach might contribute to elucidating thalamocortical projection to the minicolumns, which are subdivisions in the barrel structures with transverse diameters of 40–50 µm (Bruno et al. 2003; reviewed in Mountcastle 2003).
One unsolved issue is that we studied the thalamocortical circuit in a developmental stage. Our thalamocortical slices were obtained from young mice (postnatal days 12–16). It therefore remains unclear how the thalamocortical circuit we studied reflects mature thalamocortical circuits. We should address this issue in the future, although it is not technically easy to apply the correlation-based paradigm to adult mice (>4 wk old).
Synaptic organization of the intracortical inhibition
As shown in Fig. 6, we observed a high probability of FS RS inhibitory connections. In thalamocortical transmission, the FS RS inhibitory connections could be used for feedforward inhibition if TC fibers connect onto both the RS and FS cells, and also used for lateral inhibition if TC fibers connect onto only FS cells. It is an important issue to determine which the FS RS connections are preferentially used for. However, as shown in RS2 and FS cells of Fig. 5B1, we frequently observed the coexistence of the divergent TC RS/FS connections (TC2 Stim. and TC3 Stim. in Fig. 5B1) and the nondivergent TC FS connections (TC1 Stim. in Fig. 5B1) onto the same pairs of RS and FS cells (FS RS2 IPSPs were observed in Fig. 5B1; data not shown). Thus it is likely that individual FS RS inhibitory connections are used for both feedforward and lateral inhibition, depending on the individual TC fibers.
Figure 6B shows that the amplitudes of unitary IPSPs onto two RS cells were strongly correlated. This correlation indicates that FS cells, which generate larger IPSPs onto one RS cells, also generate larger IPSPs onto the other RS cells. Thus it is likely that the amplitudes of FS-evoked IPSPs onto RS cells are determined by presynaptic FS cells, rather than by postsynaptic RS cells. This finding functionally implies that the strength of feedforward inhibition is determined by which FS cells are activated.
We reported a high probability of detecting FS RS inhibitory connections (66.3%; Fig. 6), which is consistent with a previous study using young (2–3 wk old) mice (roughly 50%; Gabernet et al. 2005). On the other hand, the probability is lower in 2-wk-old rats (44%; Beierlein et al. 2003) and even lower in older (3–5 wk old) rats (21%; Sun et al. 2006). Thus the high connectivity in our recordings seems to be arise from both species differences and age differences.
Functional implication
A recent study using thalamocortical slices has shown that thalamic stimulation evoked highly synchronous IPSCs in adjacent spiny stellate cells in layer 4 (Sun et al. 2006). One of the likely mechanisms for the synchronous IPSCs is derived from a group of FS cells that were mutually connected with electrical and chemical synapses (Gibson et al. 1999). In the present study, we found that adjacent RS cells frequently received divergent inhibitory inputs from the same FS cell (Fig. 6). Thus it is likely that the inhibitory divergence of FS cells to RS cells may also participate in the thalamus-evoked synchronous IPSCs in RS cells.
In conclusion, we found that 1) FS cells receive highly convergent excitatory inputs from multiple TC cells and then send divergent inhibitory outputs to multiple RS cells and that 2) TC fibers generating excitatory inputs onto RS cells also generate divergent excitatory inputs onto adjacent FS cells. The former finding implies the feedforward inhibition with a low activation threshold: a relatively small number of TC cells activate action potentials in FS cells, which in turn inhibit multiple RS cells. The latter finding implies the precise feedforward inhibition: individual TC RS excitatory connections are tightly coupled to feedforward TC FS RS inhibitory connections. It will be interesting to directly demonstrate these implications, by uncovering the dynamic regulation of ascending signals in the feedforward inhibitory circuit.
|
GRANTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
Address for reprint requests and other correspondence: K. Imoto, Department of Information Physiology, National Institute for Physiological Sciences, Okazaki 444-8787, Japan (E-mail: keiji{at}nips.ac.jp)
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Agmon A and Connors BW. Correlation between intrinsic firing patterns and thalamocortical synaptic responses of neurons in mouse barrel cortex. J Neurosci 12: 319–329, 1992.[Abstract]
Amitai Y, Gibson JR, Beierlein M, Patrick SL, Ho AM, Connors BW, and Golomb D. The spatial dimensions of electrically coupled networks of interneurons in the neocortex. J Neurosci 22: 4142–4152, 2002.
Armstrong-James M and Fox K. Spatiotemporal convergence and divergence in the rat SI "barrel" cortex. J Comp Neurol 263: 265–281, 1987.[CrossRef][ISI][Medline]
Beaulieu C. Numerical data on neocortical neurons in adult rat, with specific reference to the GABA population. Brain Res 609: 284–292, 1993.[CrossRef][ISI][Medline]
Beierlein M, Gibson JR, and Connors BW. Two dynamically distinct inhibitory networks in layer 4 of the neocortex. J Neurophysiol 90: 2987–3000, 2003.
Beierlein M, Fall CP, Rinzel J, and Yuste R. Thalamocortical bursts trigger recurrent activity in neocortical networks: layer 4 as a frequency-dependent gate. J Neurosci 22: 9885–9894, 2002.
Benshalom G and White EL. Quantification of thalamocortical synapses with spiny stellate neurons in layer IV of mouse somatosensory cortex. J Comp Neurol 253: 303–314, 1986.[CrossRef][ISI][Medline]
Blitz DM and Regehr WG. Timing and specificity of feed-forward inhibition within the LGN. Neuron 45: 917–928, 2005.[CrossRef][ISI][Medline]
Bruno RM, Khatri V, Land PW, and Simons DJ. Thalamocortical angular tuning domains within individual barrels of rat somatosensory cortex. J Neurosci 23: 9565–6574, 2003.
Bruno RM and Simons DJ. Feedforward mechanisms of excitatory and inhibitory cortical receptive fields. J Neurosci 22: 10966–10975, 2002.
Castro-Alamancos MA. Dynamics of sensory thalamocortical synaptic networks during information processing states. Prog Neurobiol 74: 213–247, 2004.[CrossRef][ISI][Medline]
Deans MR, Gibson JR, Sellitto C, Connors BW, and Paul DL. Synchronous activity of inhibitory networks in neocortex requires electrical synapses containing connexin36. Neuron 31: 477–485, 2001.[CrossRef][ISI][Medline]
Gabernet L, Jadhav SP, Feldman DE, Carandini M, and Scanziani M. Somatosensory integration controlled by dynamic thalamocortical feed-forward inhibition. Neuron 48: 315–327, 2005.[CrossRef][ISI][Medline]
Gibson JR, Beierlein M, and Connors BW. Two networks of electrically coupled inhibitory neurons in neocortex. Nature 402: 75–79, 1999.[CrossRef][Medline]
Gil Z, Connors BW, and Amitai Y. Efficacy of thalamocortical and intracortical synaptic connections: quanta, innervation, and reliability. Neuron 23: 385–397, 1999.[CrossRef][ISI][Medline]
Jensen KF and Killackey HP. Terminal arbors of axons projecting to the somatosensory cortex of the adult rat. I. The normal morphology of specific thalamocortical afferents. J Neurosci 7: 3529–3543, 1987.[Abstract]
Kawaguchi Y. Physiological subgroups of nonpyramidal cells with specific morphological characteristics in layer II/III of rat frontal cortex. J Neurosci 15: 2638–2655, 1995.[Abstract]
Keller A and White EL. Synaptic organization of GABAergic neurons in the mouse SmI cortex. J Comp Neurol 262: 1–12, 1987.[CrossRef][ISI][Medline]
Killackey HP. Anatomical evidence for cortical subdivisions based on vertically discrete thalamic projections from the ventral posterior nucleus to cortical barrels in the rat. Brain Res 51: 326–331, 1973.[CrossRef][ISI][Medline]
Lin CS, Lu SM, and Schmechel DE. Glutamic acid decarboxylase immunoreactivity in layer IV of barrel cortex of rat and mouse. J Neurosci 5: 1934–1939, 1985.[Abstract]
Lübke J, Egger V, Sakmann B, and Feldmeyer D. Columnar organization of dendrites and axons of single and synaptically coupled spiny neurons in layer 4 of the rat barrel cortex. J Neurosci 20: 5300–5311, 2000.
Markram H, Lübke J, Frotscher M, Roth A, and Sakmann B. Physiology and anatomy of synaptic connections between thick tufted pyramidal neurons in the developing rat neocortex. J Physiol 500: 409–440, 1997.[CrossRef][ISI][Medline]
Moore CI and Nelson SB. Spatio-temporal subthreshold receptive fields in the vibrissa representation of rat primary somatosensory cortex. J Neurophysiol 80: 2882–2892, 1998.
Mountcastle VB. Introduction: computation in cortical columns. Cereb Cortex 13: 2–4, 2003.
Porter JT, Johnson CK, and Agmon A. Diverse types of interneurons generate thalamus-evoked feedforward inhibition in the mouse barrel cortex. J Neurosci 21: 2699–2710, 2001.
Simons DJ. Response properties of vibrissa units in rat SI somatosensory neocortex. J Neurophysiol 41: 798–820, 1978.
Simons DJ and Woolsey TA. Morphology of Golgi-Cox-impregnated barrel neurons in rat SmI cortex. J Comp Neurol 230: 119–132, 1984.[CrossRef][ISI][Medline]
Sun QQ, Huguenard JR, and Prince DA. Barrel cortex microcircuits: thalamocortical feedforward inhibition in spiny stellate cells is mediated by a small number of fast-spiking interneurons. J Neurosci 26: 1219–1230, 2006.
Swadlow HA. Influence of VPM afferents on putative inhibitory interneurons in S1 of the awake rabbit: evidence from cross-correlation, microstimulation, and latencies to peripheral sensory stimulation. J Neurophysiol 73: 1584–1599, 1995.
Swadlow HA. Fast-spike interneurons and feedforward inhibition in awake sensory neocortex. Cereb Cortex 13: 25–32, 2003.
Swadlow HA and Gusev AG. Receptive-field construction in cortical inhibitory interneurons. Nat Neurosci 5: 403–404, 2002.[CrossRef][ISI][Medline]
Waite PME. Trigeminal sensory system. In: The Rat Nervous System, edited by Paxinos G. London: Elsevier Academic Press, 2004, p. 817–851.
White EL, Benshalom G, and Hersch SM. Thalamocortical and other synapses involving nonspiny multipolar cells of mouse SmI cortex. J Comp Neurol 229: 311–320, 1984.[CrossRef][ISI][Medline]
Wong-Riley M. Changes in the visual system of monocularly sutured or enucleated cats demonstrable with cytochrome oxidase histochemistory. Brain Res 171: 11–28, 1979.[CrossRef][ISI][Medline]
Woolsey TA and van der Loos H. The structural organization of layer IV in the somatosensory region (SI) of mouse cerebral cortex. Brain Res 17: 205–242, 1970.[CrossRef][ISI][Medline]
Zhu JJ and Connors BW. Intrinsic firing patterns and whisker-evoked synaptic responses of neurons in the rat barrel cortex. J Neurophysiol 81: 1171–1183, 1999.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH |
