Les terminaisons corticothalamiques sur les neurones thalamiques réticulaires (RE) comptent pour la plupart des synapses provenant des voies afférentes sur ce noyau. Étant donné la suprématie des entrées corticales, nous avons analysé ici les caractéristiques et les mécanismes possibles sous-jacents à une composante secondaire de dépolarisations provoquées corticalement dans les neurones RE, enregistré sur des chats sous anesthésie. Une stimulation électrique des axones corticothalamiques dans la capsule interne a provoqué des potentiels postsynaptiques excitateurs (EPSPs) à latence fixe et courte qui, en augmentant l’intensité de la stimulation et à des niveaux hyperpolarisés (< -70 mV), se sont développés en décharges à seuil bas (LTSs) et en oscillations en fuseaux. Le seuil pour les oscillations en fuseaux était de 60% plus haut que ce qui est requis pour provoquer des EPSPs minimaux. Les EPSPs provoqués comportaient une composante dépolarisante secondaire qui se produisait comme un événement tout ou rien à ~5 ms après le pic de la composante initiale et était dépendante du voltage, i.e. plus présente entre –70 mV et –85 mV alors qu’elle était grandement réduite ou absente à des niveaux plus hyperpolarisés. La composante dépolarisante secondaire était sensible au QX-314 dans les micropipettes d’enregistrements. Nous suggérons que la composante secondaire des EPSPs provoqués corticalement dans les neurones RE est due à l’activation des courants-T avec une contribution probable du courant persistant Na⁺. Cette composante tardive affectait les propriétés intégratives des neurones RE, incluant leurs décharges et la sommation temporale des entrées corticales.

Corticothalamic terminals on thalamic reticular (RE) neurons account for most synapses from afferent pathways onto this nucleus and these inputs are more powerful than those from axon collaterals of thalamocortical (TC) neurons. Given the supremacy of cortical inputs, we analyzed here the characteristics and possible mechanisms underlying a secondary component of the cortically elicited depolarization in RE neurons, recorded in cats under barbiturate anesthesia. Electrical stimulation of corticothalamic axons in the internal capsule evoked fixed and short-latency excitatory postsynaptic potentials (EPSPs) that, by increasing stimulation intensity and at hyperpolarized levels (< -70 mV), developed into low-threshold spikes (LTSs) and spindle oscillations. The threshold for spindle oscillations was just 60% higher than that required for evoking minimal EPSPs. The evoked EPSPs included a secondary depolarizing component, which occurred as an all-or-none event at ~5 ms after the peak of the initial component and was voltage–dependent, i.e. most prominent between -70 mV and -85 mV, while being greatly reduced or absent at more hyperpolarized levels. The secondary depolarizing component was sensitive to QX-314 in the recording micropipette. We suggest that the secondary component of cortically evoked EPSPs in RE neurons is due to the dendritic activation of T-current, with a probable contribution of the persistent Na⁺ current. This late component affected the integrative properties of RE neurons, including their spiking output and temporal summation of incoming cortical inputs.

The thalamic reticular (RE) nucleus is entirely composed of neurons releasing γ-aminobutyric acid (GABA), which project to the dorsal thalamus and are crucially implicated in the generation of sleep spindle oscillations (Steriade et al. , 1990). The major extrinsic synaptic inputs of these neurons arise from axons of layer VI cortical pyramidal neurons and axon collaterals of thalamocortical (TC) neurons (Jones, 1985). Intrinsic synapses are both chemical (Deschênes et al. , 1985; Yen et al. , 1985) and electrical (Landisman et al. , 2002). Of all external sources of synaptic contacts, the cortex is the most important as corticothalamic terminals on RE cells in the somatosensory sector of the nucleus account for ~65-70% of synapses, whereas synapses of TC origin account for 20-25% and GABAergic synapses for 15-20% (Liu & Jones, 1999). Differences in the strength of corticothalamic connections to RE and TC cells, as expressed by three-fold larger unitary excitatory postsynaptic currents in RE neurons, are correlated with about three times higher numbers of GluR4 receptor subunits at the RE synapse (Golshani et al. , 2001). The small terminals from corticothalamic collaterals have a single small postsynaptic density, which appears to reflect the presence of a single vesicle release site, and are distributed in approximately equal numbers over both proximal and distal dendrites at RE neurons (Liu et al. , 2001).

The RE neurons display two firing modes: tonic discharges at relatively depolarized levels of membrane potential (V_m), as in brain-active behavioral states, and prolonged spike-bursts during oscillations characteristic for natural slow-wave sleep (Steriade et al. , 1986, Steriade et al. , 1990). These spike-bursts are made of fast Na⁺ action potentials that crown a low-threshold spike (LTS) generated by the transient Ca²⁺ current, I_T, which is de-inactivated by membrane hyperpolarization (Mulle et al. , 1986; Llinás & Geijo-Barrientos, 1988; Huguenard & Prince, 1992; Contreras et al. , 1993). The distal dendritic origin of RE-cells’ LTSs has been shown experimentally and in computational studies (Huguenard & Prince, 1992; Destexhe et al. , 1996). While the dendritically generated I_T in RE neurons is mainly implicated in the induction of low-frequency sleep spindle oscillations (Steriade et al. , 1987; Steriade et al. , 1990), activation of a distinct Ca²⁺ current, P/Q-type, generates high-frequency (20-80 Hz) oscillations in the dendrites of TC neurons (Pedroarena & Llinás, 1997).

In previous studies we have shown that corticothalamic stimulation elicits powerful EPSPs in RE neurons, which develop into spike-bursts and spindle oscillations in the thalamocortical network (Contreras & Steriade, 1996). The present study was undertaken to further characterize the secondary depolarizing component of cortically evoked EPSP in RE neurons (Contreras et al. , 1993) by analyzing its voltage-dependency and sensitivity to QX-314, since this component is of importance for the integrative properties of RE neurons.

Preparation

Experiments were performed on adult cats (2.5-3.5 kg), anesthetized with pentobarbital (25 mg/kg, i.p.). When the cats showed the signs of deep anaesthesia, they were paralyzed with gallamine triethiodide and artificially ventilated with control of the end-tidal CO₂ concentration at ~3.5%. Body temperature was maintained at 36-38^o C. The depth of anesthesia was continuously monitored by EEG and additional doses of anesthetic were administered at the slightest tendency toward low-voltage and fast EEG rhythms. At the end of experiments, animals were given a lethal dose of pentobarbital (50 mg/kg).

Recording and stimulation

Current-clamp intracellular recordings from the rostral and rostrolateral sector of the RE nucleus were performed using sharp electrodes, glass micropipettes (DC resistance, 30-60 MΩ). To avoid breaking of recording micropipettes, the cortex and white matter overlying the head of the caudate nucleus were removed by suction. The pipettes entered ~3 mm through the caudate nucleus to reach the RE nucleus. Pipettes were generally filled with 3 M solution of K-acetate and, in some experiments, 50 mM of QX-314 was added. The stability of intracellular recordings was ensured by cisternal drainage, bilateral pneumothorax, hip suspension, and by filling the hole over the thalamus with 4% agar solution. A high-impedance amplifier with active bridge circuitry was used to record and inject current inside the cells. Most intracellular recordings included in the database lasted for periods longer than 30 min.

Electrical stimulation of corticothalamic fibers was performed by descending one or two bipolar stimulating electrodes to the internal capsule (anterior +13, lateral +3.5, depth +1) and applying extracellular current pulses (0.2 ms, 50-600 μA, 0.5-1 Hz). In all cases, the threshold intensity for EPSP’s generation was determined. Due to the all-or-none nature of the secondary component of responses at threshold intensities, most experiments were done with intensities 10-30% over the threshold in order to study the secondary component.

Data analysis

All analyses were performed using Igor Pro 4.0 (Wavemetrics. Inc.). Values are expressed as mean ± S.D., and t-tests were used to assess statistical differences.

Intracellular recordings from RE neurons were performed in the rostral pole and rostrolateral sector of the nucleus. All cells ( n = 32) were identified by the characteristic accelerando-decelerando pattern of their spike-bursts (inset in Fig. 5.1 A ), which occurred during spindle sequences. Resting V_ms were more negative than -70 mV in all cases (-75 ± 4 mV). At such V_ms, T-current are de-inactivated (Llinás, 1988; Llinás & Geijo-Barrientos, 1988; Huguenard & Prince, 1992) and RE neurons fire bursts of action potentials during spontaneously occurring spindles (Fig. 5.1 A ).

EPSPs evoked by stimulation of corticothalamic fibers

RE neurons were activated by electrical stimulation of corticothalamic fibers in the internal capsule (Fig. 5.1, B-C ). In all cases, stimulation evoked a short-latency EPSP. Stimulation intensity was adjusted in order to obtain a minimal response. For the neuron depicted in Fig. 5.1, low intensities (≤110 μA) failed to evoke any response. Increasing stimulation intensity produced responses from 120 μA, considered the threshold. Thus, EPSPs of small amplitude (3.4 ± 1.4 mV, n = 30), as recorded at somatic level and resting V_m, and short duration (~30 ms) were elicited with threshold stimulation intensity of 120 μA (left panels in Fig. 5.1, B-C ). Such EPSPs were not followed by long-latency responses. Since there were no additional excitatory or inhibitory components, and the latency was short and fixed, these evoked EPSPs were considered to be monosynaptic. In the same cell, increasing stimulation intensity to 180 μA induced two effects: i) the short-latency EPSP gave rise to an LTS crowned by spike-bursts at the same resting V_m as that at which lower intensity stimulation evoked a simple EPSP (middle superimposition within the right column in Fig. 5.1 C ), while at more depolarized or hyperpolarized levels (upper and lower superimpositions in Fig. 5.1 C ) the EPSPs gave rise to single action potentials or displayed increased amplitude, respectively; and ii) spindle oscillation was generated in all cases (Fig. 5.1 B ). The threshold for spindle generation, a network phenomenon, was just ~60% higher than the threshold for EPSP. These results were consistently seen in all other neurons.

Progressively increasing the strength of stimulation enhanced both the amplitude and depolarization area of evoked responses (traces a-c in Fig. 5.2 A ). When stimulation intensity was strong enough, LTSs were triggered from the resting V_m (-80 mV in the neuron depicted in Fig. 5.2 A ) and were followed by either a single, delayed spike (trace d in Fig. 5.2 A ) or a burst of spikes (trace e in Fig. 5.2 A ). The relation between rising stimulation intensity and increased firing probability of RE neurons was well fitted with a sigmoid function (p_0.5 reached at 150 μA; upper plot in Fig. 5.2 A ), proving a non-linear dependence of spike probability on stimulation intensity. However, both amplitude and area of EPSP displayed a positive linear relation to stimulation intensity, since both were well fitted by a linear function (lower plot in Fig. 5.2 A ). In the majority of cases (7 out of 9 neurons), stimulation inducing spindles was also able to evoke consistent discharges crowning the initial EPSP (Fig. 5.1 A ), suggesting a critical correlation between firing of RE neurons and spindle generation.

Evoked EPSPs display a secondary depolarizing component

Most EPSPs evoked by stimulation at threshold intensity displayed a secondary depolarizing component that frequently occurred as an all-or-none event. One example is the neuron depicted in Figure 5.2 B (responses were recorded at the same V_m and evoked by stimuli with the same parameters). At the resting V_m (-76 mV) the cortically evoked EPSP displayed a short-latency, fast rising EPSP that was present in all cases (component a ), but was associated in the majority of cases (70%) with a later, depolarizing component (component b ). The latter response considerably increased (~ 75%) the depolarization area of the EPSP (from 62 to 109 mV*ms, Fig. 5.2 B ).

Effects of QX-314 on evoked EPSPs

Dendrites of most neurons are provided with a plethora of voltage-gated ionic channels (Llinás, 1975). This property allows dendrites to control synaptic amplitude and its conduction to the soma. The dendritic ionic channels may produce supralinear, sublinear or linear summation of arriving inputs. To evaluate the effect of active conductances on evoked EPSPs in RE neurons, recordings were performed in the presence of QX-314. QX-314 decreased the amplitude and depolarization area of evoked EPSPs ( n = 7). The effect of QX-314 was mainly exerted on the late component of the response (Fig. 5.3 A ). The amplitude of the first component and area of the secondary component of the EPSPs, comparing the beginning of recording (control) and after seven minutes under QX-314, were 8.1 ± 1.1 mV and 207.3 ± 23.8 mV*ms, and 7.3 ± 2.3 mV and 171.2 ± 38.6 mV*ms, respectively. QX-314 is generally used for its property as a blocker of fast and persistent Na⁺ currents (Yeh, 1978; Crill, 1996); however, it also blocks low- and high-voltage activated Ca²⁺ currents (Talbot & Sayer, 1996), K⁺ currents (Svoboda et al. , 1997; Paré & Lang, 1998), and hyperpolarization-activated currents (Perkins & Wong, 1995).

Voltage dependence of the secondary depolarizing component and the effect of QX-314

In a set of neurons ( n = 7), evoked EPSPs were recorded at rest and at different values of V_m, obtained by current injection through the pipette (Fig. 5.3 B ). V_m depolarization from rest (-78 mV, second trace in left panel of Fig. 5.3 B ) to –73 mV increased the depolarization area of EPSPs, mainly by enhancing the secondary depolarizing component, whereas the amplitude of the early component remained nearly invariable. At more negative values of the V_m (-88 mV to –106 mV), the presence of the secondary depolarizing component was markedly diminished, while the amplitude of the first response remained virtually the same (left panel in Fig. 5.3 B ). Measurements of both early and late components showed that a linear relation was kept in both cases; while the amplitude of the early component remained almost invariable, the area of the secondary depolarizing component increased linearly with depolarization (Fig. 5.3 C ).

QX-314 decreased both the amplitude and area of cortically evoked EPSPs in a voltage dependent way ( n = 4). The RE neuron depicted in Fig. 5.3 B (right panel) was recorded during 30 min in QX-314 (50 mM). After an initial period of 2-3 min, full action potentials decreased in both amplitude and incidence, showing an effect of QX-314 on fast Na⁺ currents (not shown). At rest (-83 mV), cortically evoked EPSPs showed rather small amplitude ( ~ 5 mV), but no secondary depolarizing component was evident (right panel in Fig. 5.3 B ). Contrary to control recordings (with K⁺-acetate), EPSPs recorded under QX-314 displayed decreased amplitude and area of depolarization as the V_m became more positive.

Figure 5.3 C presents the pooled results for various neurons recorded under either control conditions (K⁺-acetate, n = 7) or QX-314 ( n = 4). For all cases recorded under QX-314, a voltage dependent decrease in both amplitude and area of evoked EPSPs was found (Fig. 5.3 C ). QX-314 started to exert its effect at about -80 mV, a value close to the resting V_m for RE neurons, and the effect increased as V_m depolarized.

Properties of suprathreshold synaptic responses

Since in most cases the resting V_m was quite hyperpolarized (around -75 mV or even more negative), EPSPs were generally subthreshold for spike generation. Increasing stimulation intensity frequently led to spike generation, but those usually crowned the LTSs (see Fig. 5.2 A ). Depolarizing the V_m by steady current injection through the pipette was another way to elicit suprathreshold responses. In such cases, the early and late components of evoked EPSPs showed differential contributions to neuronal output ( n = 11). The RE neuron depicted in Fig. 5.4 was recorded at a resting V_m (-80 mV) at which evoked EPSPs were not able to trigger action potentials (not shown). However, as the neuron was depolarized, spikes were triggered in response to cortically evoked EPSPs. At -73 mV, spikes were triggered by the secondary depolarizing component ( b ), while the early component ( a ) remained subthreshold (Fig. 5.4 A ). Only starting at -65 mV was the early component able to trigger spikes (dotted line). The probability of spike generation was plotted for both early and late components of the response as a function of V_m, and both were well fitted with a sigmoid function (Fig. 5.4 B ). The secondary component of the response was able to trigger full action potentials at more hyperpolarized V_m values than the early component. In fact, the V_0.5 for the early component was about -64 mV, and about -72 mV for the secondary component (Fig. 5.4 B ). Thus, the secondary depolarizing component of evoked EPSPs was able to boost the neuronal output of RE neurons at a more negative V_m.

The boosting properties of the secondary component were correlated with a faster rising phase as V_m was depolarized (Fig. 5.4 A ). Plotting the slow rising slope of the secondary depolarizing component against the holding V_m showed an increasing slope with depolarization. Such relation was well fitted with a single exponential function, which describes the dynamics of the process (Fig. 5.4 B ). In contrast, the early component of the EPSP presented a very fast, though only slightly voltage dependent rising slope (Fig. 5.4 B ).

The variable rising slope of the secondary component was expressed as a wide range (3-25 ms) of variable latencies in spike generation, standing in contrast to a short, fixed latency for the early component (Fig. 5.4 A ). Plotting the spike latency as a function of V_m showed an inverse relation compared to the rising slope for the secondary depolarizing component, meaning a progressive decrease as the neuronal membrane was depolarized (Fig. 5.4 B ). Such relation was well fitted with a single, decaying exponential, while the early component displayed an almost independent relation of spike-latency to V_m (Fig. 5.4 B ).

These results show that the early and late components of evoked EPSPs have differential contributions in the neuronal output. The early component generates stereotyped spikes at short, fixed latencies, at relatively depolarized values (around -65 mV), while the secondary depolarizing component triggers spikes with widely variable latencies at more hyperpolarized values (around -72 mV).

Temporal summation of evoked EPSPs

The temporal summation of cortical inputs was assessed by high frequency stimulation (50-100 Hz) of corticothalamic fibers in the internal capsule ( n = 8). The temporal summation of EPSPs was high and voltage-dependent. Figure 5.5 shows a RE neuron that was stimulated with 5 pulses at 10-ms time- intervals, at different V_ms. Only at V_ms close to -80 mV or more positive, was the summation effect observed (Fig. 5.5 A ). At rest (-82 mV), slight summation was detected, which increased with V_m depolarization. Plotting the ratio between the last (fifth) and first EPSP against the holding V_m, showed a maximal (200%) summation (Fig. 5.5 B ). Around -76 mV the responses to the pulse-train ended with a spike in the last stimulus (Fig. 5.5 A ). Increasing the membrane depolarization shifted the initiation of spikes to earlier EPSPs in the train in a linear fashion. Thus, in this example, at -76 mV only the last EPSP was able to trigger a spike, while at -53 mV the first EPSP triggers a full action potential (Fig. 5.5 B ).

Our results indicate that synaptic connections made by corticothalamic fibers onto RE neurons display some remarkable features: (a) the stimulation intensity threshold for spindle generation, a network phenomenon, was just ~60% higher than the threshold for short-latency and small-amplitude, monosynaptic EPSP; (b) evoked EPSPs frequently displayed a secondary depolarizing component, which was voltage dependent and sensitive to intracellular application of QX-314; (c) both rising phase and amplitude of early components of evoked responses were also affected by QX-314; and (d) the secondary depolarizing component of evoked EPSPs influenced the neuronal output and temporal summation of cortical inputs.

The neocortex and thalamus engage in a continuous dialogue that is maintained by corticothalamic and thalamocortical fibers, which connect them in a recurrent circuit. Although the stimulating electrode within the internal capsule likely activated not only corticothalamic fibers but also antidromically invaded some thalamocortical axons, with possibly axon reflex excitation of RE neurons, the number of corticothalamic axons is one order of magnitude higher than that of thalamocortical ones. Thus, we consider that the evoked EPSPs were mainly due to the activation of corticothalamic fibers and the contribution of thalamocortical fibers was minimal.

In the present experiments, stimulation was adjusted for minimal intensity and we could determine that the threshold for evoking spindle oscillations was only 40% higher than for EPSPs. Increasing stimulation intensity showed that both the amplitude and depolarization area of the response rose in a linear way. However, the probability of discharge seemed to be governed by a sigmoid function. Spindle oscillations are known to be a network phenomenon initiated in the RE nucleus and maintained by the recurrent intrathalamic circuit (Steriade et al. , 1993). Thus, firing in RE neurons is necessary to generate the oscillation, which explains the close relation between the threshold for EPSP generation and spindle activation. At hyperpolarized resting V_m (around -75 mV), as displayed by RE neurons recorded in present experiments, the T-current is fully de-inactivated and therefore LTSs can be generated (Huguenard, 1996). Actually, all recorded neurons displayed spontaneous bursting behavior during spindle oscillations. The graded nature of LTSs in RE neurons was well characterized in vivo (Contreras et al. , 1993) and depends on the distal dendritic localization of T-channels (Huguenard & Prince, 1992; Destexhe et al. , 1996) coupled with the constant synaptic bombardment of dendritic arbors by network activity (Contreras et al. , 1993).

We detected a secondary depolarizing component in evoked EPSPs of RE neurons, as was also previously reported (Contreras et al. , 1993). In the present study, this component occasionally occurred as an all-or-none event, appearing ~5 ms after the peak of the response, in cases of small amplitude EPSPs. Higher amplitude EPSPs presented always a secondary depolarizing component. The secondary depolarizing component of the EPSP was voltage dependent, as it was present at V_ms as hyperpolarized as -100 mV, linearly increased with depolarization, and gave rise to action potentials as the V_m was close to firing threshold. Besides its voltage dependency, the secondary depolarizing component was sensitive to QX-314 in the recording pipette. The effect of QX-314 was also voltage dependent since it did not affect responses at voltages more negative than -80 mV. QX-314 is a blocker of many conductances (see RESULTS ). However, given the hyperpolarized V_m at which QX-314 exerted its effect, the secondary depolarizing component was probably due to the dendritic activation of T-current. The secondary component of evoked responses seems to be important for the integrative properties of RE neurons since it modulated both the neuronal output and its precise timing. Indeed, the secondary component was able to boost spike generation in RE neurons at V_ms more negative by 5-10 mV and to generate single or multiple spikes within a variable time window, as compared to stereotyped spikes generated by the peak of the early component of EPSPs when it reached the threshold. Consistent with an integrative role of the secondary depolarization, high-frequency stimulation induced temporal summation of evoked responses in a voltage dependent manner. These properties are related to the previously characterized gradual nature of bursting responses in RE cells (Contreras et al. , 1993).

The presence and contribution of persistent Na⁺ currents cannot be discarded since its range of activation is close to the threshold for spike generation (5-10 mV below) and it is also blocked by QX-314 (Crill, 1996). Besides, studies in inhibitory neurons from the hippocampus have demonstrated the presence and importance of active dendritic Na⁺ conductances (Martina et al. , 2000). It is thus possible that a joint contribution of Ca²⁺ and Na⁺ dendritic currents acts in RE neurons for integration of cortical inputs.

Another effect of QX-314 on evoked responses was the two- to three-fold decrease in the rising slope (not shown). The rising phase of any active response depends on both its location in the dendritic tree and the active conductances that regulate its conduction to the soma (Magee, 2000). In the present experiments, electrical stimulation of corticothalamic fibers frequently elicited fast-rising responses (5-10 mV/ms slope) in RE neurons. Recent experiments have shown that cortically evoked EPSPs are mainly AMPA-receptor-dependent, with a negligible element of NMDA-receptor response, which is virtually absent at hyperpolarized levels of V_m (-70 mV) (Gentet & Ulrich, 2004; see also Pedroarena & Llinás, 1997). Even though the effect of QX-314 on the rising phase was not voltage dependent, a clear reduction was seen at different voltages in all cases, suggesting an additive contribution of active conductances to the AMPA component in the response.

In short, then, the voltage dependent dendritic channel distribution in RE neurons provides them with important properties for integration of cortical inputs. We suggest that, in comparison with the effects these inputs exert on TC neurons, at which level dendritic currents generate high-frequency rhythms (Pedroarena & Llinás, 1997) that define brain alertness, the parallel activation of RE-cells’ dendrites by synchronized cortical volleys produces low-frequency oscillations (Steriade et al. , 1990; Contreras et al. , 1993; Contreras & Steriade, 1996) that characterize the disconnected state of slow-wave sleep. Moreover, during cortically generated spike-wave seizures, the powerful activation of GABAergic RE neurons leads to the inhibition of TC neurons (Steriade, 2003), their steady hyperpolarization with increased membrane conductance and, consequently, the obliteration of incoming messages en route to the cortex.

Contreras, D. & Steriade, M. (1996) Spindle oscillation in cats: the role of corticothalamic feedback in a thalamically generated rhythm. J. Physiol. , 490 , 159-179.

Crill, W.E. (1996) Persistent sodium current in mammalian central neurons. Annu. Rev. Physiol. , 58 , 349-362.

Deschênes, M., Madariaga-Domich, A. & Steriade, M. (1985) Dendrodendritic synapses in the cat reticularis thalami nucleus: a structural basis for thalamic spindle synchronization. Brain Res. , 334 , 165-168.

Destexhe, A., Contreras, D., Steriade, M., Sejnowski, T.J. & Huguenard, J.R. (1996) In vivo, in vitro, and computational analysis of dendritic calcium currents in thalamic reticular neurons. J. Neurosci. , 16 , 169-185.

Gentet, L.J. & Ulrich, D. (2004) Electrophysiological characterization of synaptic connections between layer VI cortical cells and neurons of the nucleus reticularis thalami in juvenile rats. Eur. J. Neurosci. , 19 , 625-633.

Golshani, P., Liu, X.B. & Jones, E.G. (2001) Differences in quantal amplitude reflect GluR4- subunit number at corticothalamic synapses on two populations of thalamic neurons. Proc. Natl. Acad. Sci. U S A , 98 , 4172-4177.

Huguenard, J.R. (1996) Low-threshold calcium currents in central nervous system neurons. Annu. Rev. Physiol. , 58 , 329-348.

Huguenard, J.R. & Prince, D.A. (1992) A novel T-type current underlies prolonged Ca(2+)-dependent burst firing in GABAergic neurons of rat thalamic reticular nucleus. J. Neurosci. , 12 , 3804-3817.

Jones, E.G. (1985) The Thalamus . Plenum, New York.

Landisman, C.E., Long, M.A., Beierlein, M., Deans, M.R., Paul, D.L. & Connors, B.W. (2002) Electrical synapses in the thalamic reticular nucleus. J. Neurosci. , 22 , 1002-1009.

Liu, X.B., Bolea, S., Golshani, P. & Jones, E.G. (2001) Differentiation of corticothalamic and thalamocortical collateral synapses on mouse reticular nucleus neurons by EPSC amplitude and AMPA-receptor subunit composition. Thalamus & Related Systems , 1 , 15.

Liu, X.B. & Jones, E.G. (1999) Predominance of corticothalamic synaptic inputs to thalamic reticular nucleus neurons in the rat. J. Comp. Neurol. , 414 , 67-79.

Llinás, R. (1975) Electroresponsive properties of dendrites in central neurons. Adv. Neurol. , 12 , 1-13.

Llinás, R. (1988) The intrinsic electrophysiological properties of mammalian neurons: insights into central nervous system function. Science , 242 , 1654-1664.

Llinás, R. & Geijo-Barrientos, E. (1988) Cellular Thalamic Mechanisms . Elsevier, Amsterdam.

Magee, J.C. (2000) Dendritic integration of excitatory synaptic input. Nat. Rev. Neurosci. , 1 , 181-190.

Martina, M., Vida, I. & Jonas, P. (2000) Distal initiation and active propagation of action potentials in interneuron dendrites. Science , 287 , 295-300.

Mulle, C., Madariaga, A. & Deschênes, M. (1986) Morphology and electrophysiological properties of reticularis thalami neurons in cat: in vivo study of a thalamic pacemaker. J. Neurosci. , 6 , 2134-2145.

Paré, D. & Lang, E.J. (1998) Calcium electrogenesis in neocortical pyramidal neurons in vivo. Eur. J. Neurosci. , 10 , 3164-3170.

Pedroarena, C. & Llinás, R. (1997) Dendritic calcium conductances generate high-frequency oscillation in thalamocortical neurons. Proc. Natl. Acad. Sci. U S A , 94 , 724-728.

Perkins, K.L. & Wong, R.K. (1995) Intracellular QX-314 blocks the hyperpolarization-activated inward current Iq in hippocampal CA1 pyramidal cells. J. Neurophysiol. , 73 , 911-915.

Steriade, M. (2003) Neuronal Substrates of Sleep and Epilepsy . Cambridge University Press, Cambridge, UK.

Steriade, M., Domich, L. & Oakson, G. (1986) Reticularis thalami neurons revisited: activity changes during shifts in states of vigilance. J. Neurosci. , 6 , 68-81.

Steriade, M., Domich, L., Oakson, G. & Deschênes, M. (1987) The deafferented reticular thalamic nucleus generates spindle rhythmicity. J. Neurophysiol. , 57 , 260-273.

Steriade, M., Jones, E.G. & Llinás, R. (1990) Thalamic oscillations and signaling. Wiley-Interscience, New York.

Steriade, M., McCormick, D.A. & Sejnowski, T.J. (1993) Thalamocortical oscillations in the sleeping and aroused brain. Science , 262 , 679-685.

Svoboda, K., Denk, W., Kleinfeld, D. & Tank, D.W. (1997) In vivo dendritic calcium dynamics in neocortical pyramidal neurons. Nature , 385 , 161-165.

Talbot, M.J. & Sayer, R.J. (1996) Intracellular QX-314 inhibits calcium currents in hippocampal CA1 pyramidal neurons. J. Neurophysiol. , 76 , 2120-2124.

Yeh, J.Z. (1978) Sodium inactivation mechanism modulates QX-314 block of sodium channels in squid axons. Biophys. J. , 24 , 569-574.

Yen, C.T., Conley, M., Hendry, S.H. & Jones, E.G. (1985) The morphology of physiologically identified GABAergic neurons in the somatic sensory part of the thalamic reticular nucleus in the cat. J. Neurosci. , 5 , 2254-2268.