A mean-field approach to the dynamics of networks of complex neurons, from nonlinear Integrate-and-Fire to Hodgkin-Huxley models.

We present a mean-field formalism able to predict the collective dynamics of large networks of conductance-based interacting spiking neurons. We apply this formalism to several neuronal models, from the simplest Adaptive Exponential Integrate-and-Fire model to the more complex Hodgkin-Huxley and Morris-Lecar models. We show that the resulting mean-field models are capable of predicting the correct spontaneous activity of both excitatory and inhibitory neurons in asynchronous irregular regimes, typical of cortical dynamics. Moreover, it is possible to quantitatively predict the population response to external stimuli in the form of external spike trains. This mean-field formalism therefore provides a paradigm to bridge the scale between population dynamics and the microscopic complexity of the individual cells physiology.NEW & NOTEWORTHY Population models are a powerful mathematical tool to study the dynamics of neuronal networks and to simulate the brain at macroscopic scales. We present a mean-field model capable of quantitatively predicting the temporal dynamics of a network of complex spiking neuronal models, from Integrate-and-Fire to Hodgkin-Huxley, thus linking population models to neurons electrophysiology. This opens a perspective on generating biologically realistic mean-field models from electrophysiological recordings.

Heterogeneous firing rate response of mouse layer V pyramidal neurons in the fluctuation-driven regime.

KEY POINTS: We recreated in vitro the fluctuation-driven regime observed at the soma during asynchronous network activity in vivo and we studied the firing rate response as a function of the properties of the membrane potential fluctuations. We provide a simple analytical template that captures the firing response of both pyramidal neurons and various theoretical models. We found a strong heterogeneity in the firing rate response of layer V pyramidal neurons: in particular, individual neurons differ not only in their mean excitability level, but also in their sensitivity to fluctuations. Theoretical modelling suggest that this observed heterogeneity might arise from various expression levels of the following biophysical properties: sodium inactivation, density of sodium channels and spike frequency adaptation. ABSTRACT: Characterizing the input-output properties of neocortical neurons is of crucial importance for understanding the properties emerging at the network level. In the regime of low-rate irregular firing (such as in the awake state), determining those properties for neocortical cells remains, however, both experimentally and theoretically challenging. Here, we studied this problem using a combination of theoretical modelling and in vitro experiments. We first identified, theoretically, three somatic variables that describe the dynamical state at the soma in this fluctuation-driven regime: the mean, standard deviation and time constant of the membrane potential fluctuations. Next, we characterized the firing rate response of individual layer V pyramidal cells in this three-dimensional space by means of perforated-patch recordings and dynamic clamp in the visual cortex of juvenile mice in vitro. We found that individual neurons strongly differ not only in terms of their excitability, but also, and unexpectedly, in their sensitivities to fluctuations. Finally, using theoretical modelling, we attempted to reproduce these results. The model predicts that heterogeneous levels of biophysical properties such as sodium inactivation, sharpness of sodium activation and spike frequency adaptation account for the observed diversity of firing rate responses. Because the firing rate response will determine population rate dynamics during asynchronous neocortical activity, our results show that cortical populations are functionally strongly inhomogeneous in young mouse visual cortex, which should have important consequences on the strategies of cortical computation at early stages of sensory processing.

Extracting synaptic conductances from single membrane potential traces.

In awake animals, the activity of the cerebral cortex is highly complex, with neurons firing irregularly with apparent Poisson statistics. One way to characterize this complexity is to take advantage of the high interconnectivity of cerebral cortex and use intracellular recordings of cortical neurons, which contain information about the activity of thousands of other cortical neurons. Identifying the membrane potential (Vm) to a stochastic process enables the extraction of important statistical signatures of this complex synaptic activity. Typically, one estimates the total synaptic conductances (excitatory and inhibitory) but this type of estimation requires at least two Vm levels and therefore cannot be applied to single Vm traces. We propose here a method to extract excitatory and inhibitory conductances (mean and variance) from single Vm traces. This "VmT method" estimates conductance parameters using maximum likelihood criteria, under the assumption that synaptic conductances are described by gaussian stochastic processes and are integrated by a passive leaky membrane. The method is illustrated using models and is tested on guinea-pig visual cortex neurons in vitro using dynamic-clamp experiments. The VmT method holds promises for extracting conductances from single-trial measurements, which has a high potential for in vivo applications.

Control of spatiotemporal coherence of a thalamic oscillation by corticothalamic feedback.

The mammalian thalamus is the gateway to the cortex for most sensory modalities. Nearly all thalamic nuclei also receive massive feedback projections from the cortical region to which they project. In this study, the spatiotemporal properties of synchronized thalamic spindle oscillations (7 to 14 hertz) were investigated in barbiturate-anesthetized cats, before and after removal of the cortex. After complete ipsilateral decortication, the long-range synchronization of thalamic spindles in the intact cortex hemisphere changed into disorganized patterns with low spatiotemporal coherence. Local thalamic synchrony was still present, as demonstrated by dual intracellular recordings from nearby neurons. In the cortex, synchrony was insensitive to the disruption of horizontal intracortical connections. These results indicate that the global coherence of thalamic oscillations is determined by corticothalamic projections.

Model of low-pass filtering of local field potentials in brain tissue.

Local field potentials (LFPs) are routinely measured experimentally in brain tissue, and exhibit strong low-pass frequency filtering properties, with high frequencies (such as action potentials) being visible only at very short distances (approximately 10 microm) from the recording electrode. Understanding this filtering is crucial to relate LFP signals with neuronal activity, but not much is known about the exact mechanisms underlying this low-pass filtering. In this paper, we investigate a possible biophysical mechanism for the low-pass filtering properties of LFPs. We investigate the propagation of electric fields and its frequency dependence close to the current source, i.e., at length scales in the order of average interneuronal distances. We take into account the presence of a high density of cellular membranes around current sources, such as glial cells. By considering them as passive cells, we show that under the influence of the electric source field, they respond by polarization. Because of the finite velocity of ionic charge movements, this polarization will not be instantaneous. Consequently, the induced electric field will be frequency-dependent, and much reduced for high frequencies. Our model establishes that this situation is analogous to an equivalent RC circuit, or better yet a system of coupled RC circuits. We present a number of numerical simulations of an induced electric field for biologically realistic values of parameters, and show the frequency filtering effect as well as the attenuation of extracellular potentials with distance. We suggest that induced electric fields in passive cells surrounding neurons are the physical origin of frequency filtering properties of LFPs. Experimentally testable predictions are provided allowing us to verify the validity of this model.

Cortical feedback controls the frequency and synchrony of oscillations in the visual thalamus.

Thalamic circuits have an intrinsic capacity to generate state-dependent oscillations of different frequency and degrees of synchrony, but little is known of how synchronized oscillation is controlled in the intact brain or what function it may serve. The influence of cortical feedback was examined using slice preparations of the visual thalamus and computational models. Cortical feedback was mimicked by stimulating corticothalamic axons, triggered by the activity of relay neurons. This artificially coupled network had the capacity to self-organize and to generate qualitatively different rhythmical activities according to the strength of corticothalamic feedback stimuli. Weak feedback (one to three shocks at 100-150 Hz) phase-locked the spontaneous spindle oscillations (6-10 Hz) in geniculate and perigeniculate nuclei. However, strong feedback (four to eight shocks at 100-150 Hz) led to a more synchronized oscillation, slower in frequency (2-4 Hz) and dependent on GABA(B) receptors. This increase in synchrony was essentially attributable to a redistribution of the timing of action potential generation in lateral geniculate nucleus cells, resulting in an increased output of relay cells toward the cortex. Corticothalamic feedback is thus capable of inducing highly synchronous slow oscillations in physiologically intact thalamic circuits. This modulation may have implications for a better understanding of the descending control of thalamic nuclei by the cortex, and the genesis of pathological rhythmical activity, such as absence seizures.

Spatiotemporal analysis of local field potentials and unit discharges in cat cerebral cortex during natural wake and sleep states.

The electroencephalogram displays various oscillation patterns during wake and sleep states, but their spatiotemporal distribution is not completely known. Local field potentials (LFPs) and multiunits were recorded simultaneously in the cerebral cortex (areas 5-7) of naturally sleeping and awake cats. Slow-wave sleep (SWS) was characterized by oscillations in the slow (<1 Hz) and delta (1-4 Hz) frequency range. The high-amplitude slow-wave complexes consisted in a positivity of depth LFP, associated with neuronal silence, followed by a sharp LFP negativity, correlated with an increase of firing. This pattern was of remarkable spatiotemporal coherence, because silences and increased firing occurred simultaneously in units recorded within a 7 mm distance in the cortex. During wake and rapid-eye-movement (REM) sleep, single units fired tonically, whereas LFPs displayed low-amplitude fast activities with increased power in fast frequencies (15-75 Hz). In contrast with the widespread synchronization during SWS, fast oscillations during REM and wake periods were synchronized only within neighboring electrodes and small time windows (100-500 msec). This local synchrony occurred in an apparent irregular manner, both spatially and temporally. Brief periods (<1 sec) of fast oscillations were also present during SWS in between slow-wave complexes. During these brief periods, the spatial and temporal coherence, as well as the relation between units and LFPs, was identical to that of fast oscillations of wake or REM sleep. These results show that natural SWS in cats is characterized by slow-wave complexes, synchronized over large cortical territories, interleaved with brief periods of fast oscillations, characterized by local synchrony, and of characteristics similar to that of the sustained fast oscillations of activated states.

The discharge variability of neocortical neurons during high-conductance states.

In vivo recordings have shown that the discharge of cortical neurons is often highly variable and can have statistics similar to a Poisson process with a coefficient of variation around unity. To investigate the determinants of this high variability, we analyzed the spontaneous discharge of Hodgkin-Huxley type models of cortical neurons, in which in vivo-like synaptic background activity was modeled by random release events at excitatory and inhibitory synapses. By using compartmental models with active dendrites, or single compartment models with fluctuating conductances and fluctuating currents, we found that a high discharge variability was always paralleled with a high-conductance state, while some active and passive cellular properties had only a minor impact. Furthermore, a balance between excitation and inhibition was not a necessary condition for high discharge variability. We conclude that the fluctuating high-conductance state caused by the ongoing activity in the cortical network in vivo may be viewed as a natural determinant of the highly variable discharges of these neurons.

Spindle oscillations during cortical spreading depression in naturally sleeping cats.

Spindling activity characterizes the EEG of animals and humans in the early stages of resting sleep. Spindles are defined as waxing and waning rhythmic waves at 7-14 Hz that recur periodically every 3-10 s. Spindling originates in the thalamus, but a role for the cerebral cortex in triggering and synchronizing thalamic spindles was shown by stimulation of the contralateral cortex avoiding antidromic activation of thalamocortical axons and by diminished coherency of thalamic spindles after hemidecortication. Spontaneous spindles under barbiturate anesthesia are waxing and waning but under ketamine-xylazine anesthesia or when evoked by strong stimuli spindle waves are almost exclusively waning, i.e. they start with maximum amplitude and then decrease progressively. Waxing and waning of spindles has been ascribed to progressive entrainment of units into the oscillation followed by a progressive desynchronization. Therefore, exclusively waning spindles would be produced by an initial high synchrony in the corticothalamic network. Such a situation is observable upon strong stimulation or, spontaneously, when spindles are paced by the slow cortical oscillation and preceded by a strong corticothalamic drive. We have conducted experiments in naturally sleeping cats to verify the occurrence of two patterns of spindle oscillations and to test the role of the cortex in synchronizing and shaping spindles. We have found that indeed two types of spindles (waxing and waning or mostly waning) occur in naturally sleeping animals. We also demonstrate that during cortical spreading depression spindles are less synchronous and only of the waxing and waning type. As cortical activity recovers, waning spindles reappear and are preceded by electroencephalogram deflections which are related to corticothalamic depolarizing inputs. Our results strongly support the hypothesis of the role of the cerebral cortex in shaping and synchronizing thalamically generated spindles.

Dual intracellular recordings and computational models of slow inhibitory postsynaptic potentials in rat neocortical and hippocampal slices.

Dual intracellular recordings in slices of adult rat neocortex and hippocampus investigated slow, putative GABA(B) receptor-mediated inhibitory postsynaptic potentials. In most pairs tested in which the interneuron elicited a fast inhibitory postsynaptic potential in the pyramid, this GABA(A) receptor mediated inhibitory postsynaptic potential was entirely blocked by bicuculline or picrotoxin (3:3 in neocortex, 6:8 in CA1, all CA1 basket cells), even when high-frequency presynaptic spike trains were elicited. However, in three of 85 neocortical paired recordings involving an interneuron, although no discernible response was elicited by single presynaptic interneuronal spikes, a long latency (> or =20 ms) inhibitory postsynaptic potential was elicited by a train of > or =3 spikes at frequencies > or =50-100 Hz. This slow inhibitory postsynaptic potential was insensitive to bicuculline (one pair tested). In neocortex, slow inhibitory postsynaptic potential duration reached a maximum of 200 ms even with prolonged presynaptic spike trains. In contrast, summing fast, GABA(A) inhibitory postsynaptic potentials, elicited by spike trains, lasted as long as the train. Between four and 10 presynaptic spikes, mean peak slow inhibitory postsynaptic potential amplitude increased sharply to 0.38, 2.6 and 2.9 mV, respectively, in the three neocortical pairs (membrane potential -60 to -65 mV). Thereafter increases in spike number had little additional effect on amplitude. In two of eight pairs in CA1, one involving a presynaptic basket cell and the other a putative bistratified interneuron, the fast inhibitory postsynaptic potential was blocked by bicuculline revealing a slow inhibitory postsynaptic potential that was greatly reduced by 100 microM CGP 35348 (basket cell pair). The sensitivity of this slow inhibitory postsynaptic potential to spike number was similar to that of neocortical 'pure' slow inhibitory postsynaptic potentials, but was of longer duration, its plateau phase outlasting 200 ms spike trains and its maximum duration exceeding 400 ms. Computational models of GABA release, diffusion and uptake suggested that extracellular accumulation of GABA cannot alone account for the non-linear relationship between spike number and inhibitory postsynaptic potential amplitude. However, cooperativity in the kinetics of GABA(B) transduction mechanisms provided non-linear relations similar to experimental data. Different kinetic models were considered for how G-proteins activate K+ channels, including allosteric models. For all models, the best fit to experimental data was obtained with four G-protein binding sites on the K+ channels, consistent with a tetrameric structure for the K+ channels associated with GABA(B) receptors. Thus some inhibitory connections in neocortex and hippocampus appear mediated solely by fast GABA(A) receptors, while others appear mediated solely by slow, non-ionotropic, possibly GABA(B) receptors. In addition, some inhibitory postsynaptic potentials arising in proximal portions of CA1 pyramidal cells are mediated by both GABA(A) and GABA(B) receptors. Our data indicate that the GABA released by a single interneuron can saturate the GABA(B) receptor mechanism(s) accessible to it and that 'spillover' to extrasynaptic sites need not necessarily be proposed to explain these slow inhibitory postsynaptic potential properties.

Cortically-induced coherence of a thalamic-generated oscillation.

Oscillatory patterns in neocortical electrical activity show various degrees of large-scale synchrony depending on experimental conditions, but the exact mechanisms underlying these variations of coherence are not known. Analysis of multisite local field potentials revealed that the coherence of spindle oscillations varied during different states. During natural sleep, the coherence was remarkably high over cortical distances of several millimeters, but could be disrupted by artificial cortical depression, similar to the effect of barbiturates. Possible mechanisms for these variations of coherence were investigated by computational models of interacting cortical and thalamic neurons, including their intrinsic firing patterns and various synaptic receptors present in the circuitry. The model indicates that modulation of the excitability of the cortex can affect spatiotemporal coherence with no change in the thalamus. The highest level of coherence was obtained by enhancing the excitability of cortical pyramidal cells, simulating the action of neuromodulators such as acetylcholine and noradrenaline. The underlying mechanism was due to cortex-thalamus-cortex loops in which a more excitable cortical network generated a more powerful and coherent feedback onto the thalamus, resulting in highly coherent oscillations, similar to the properties measured during natural sleep. In conclusion, these experiments and models are compatible with a powerful role for the cortex in triggering and synchronizing oscillations generated in the thalamus, through corticothalamic feedback projections. The model suggests that intracortical mechanisms may be responsible for synchronizing oscillations over cortical distances of several millimeters through cortex-thalamus-cortex loops, thus providing a possible cellular mechanism to explain the genesis of large-scale coherent oscillations in the thalamocortical system.

Inhibitory control of somatodendritic interactions underlying action potentials in neocortical pyramidal neurons in vivo: an intracellular and computational study.

The effect of synaptic inputs on somatodendritic interactions during action potentials was investigated, in the cat, using in vivo intracellular recording and computational models of neocortical pyramidal cells. An array of 10 microelectrodes, each ending at a different cortical depth, was used to preferentially evoke synaptic inputs to different somatodendritic regions. Relative to action potentials evoked by current injection, spikes elicited by cortical microstimuli were reduced in amplitude and duration, with stimuli delivered at proximal (somatic) and distal (dendritic) levels evoking the largest and smallest decrements, respectively. When the inhibitory postsynaptic potential reversal was shifted to around -50 mV by recording with KCl pipettes, synaptically-evoked spikes were significantly less reduced than with potassium acetate or cesium acetate pipettes, suggesting that spike decrements are not only due to a shunt, but also to voltage-dependent effects. Computational models of neocortical pyramidal cells were built based on available data on the distribution of active currents and synaptic inputs in the soma and dendrites. The distribution of synapses activated by extracellular stimulation was estimated by matching the model to experimental recordings of postsynaptic potentials evoked at different depths. The model successfully reproduced the progressive spike amplitude reduction as a function of stimulation depth, as well as the effects of chloride and cesium. The model revealed that somatic spikes contain an important contribution from proximal dendritic sodium currents up to approximately 100 microm and approximately 300 microm from the soma under control and cesium conditions, respectively. Proximal inhibitory postsynaptic potentials can present this dendritic participation thus reducing the spike amplitude at the soma. The model suggests that the somatic spike amplitude and shape can be used as a "window" to infer the electrical participation of proximal dendrites. Thus, our results suggest that inhibitory postsynaptic potentials can control the participation of proximal dendrites in somatic sodium spikes.

Control of slow oscillations in the thalamocortical neuron: a computer model.

We investigated computer models of a single thalamocortical neuron to assess the interaction of intrinsic voltage-sensitive channels and cortical synaptic input in producing the range of oscillation frequencies observed in these cells in vivo. A morphologically detailed model with Hodgkin-Huxley-like ion channels demonstrated that intrinsic properties would be sufficient to readily produce 3 to 6 Hz oscillations. Hyperpolarization of the model cell reduced its oscillation frequency monotonically whether through current injection or modulation of a potassium conductance, simulating the response to a neuromodulatory input. We performed detailed analysis of highly reduced models to determine the mechanism of this frequency control. The interburst interval was controlled by two different mechanisms depending on whether or not the pacemaker current, IH, was present. In the absence of IH, depolarization during the interburst interval occurred at the same rate with different current injections. The voltage difference from the nadir to threshold for the low-threshold calcium current, IT, determined the interburst interval. In contrast, with IH present, the rate of depolarization depended on injected current. With the full model, simulated repetitive cortical synaptic input entrained oscillations up to approximately double the natural frequency. Cortical input readily produced phase resetting as well. Our findings suggest that neither ascending brainstem control altering underlying hyperpolarization, nor descending drive by repetitive cortical inputs, would alone be sufficient to produce the range of oscillation frequencies seen in thalamocortical neurons. Instead, intrinsic neuronal mechanisms would dominate for generating the delta range (0.5-4 Hz) oscillations seen during slow wave sleep, whereas synaptic interactions with cortex and the thalamic reticular nucleus would be required for faster oscillations in the frequency range of spindling (7-14 Hz).

Synaptic background noise controls the input/output characteristics of single cells in an in vitro model of in vivo activity.

In vivo, in vitro and computational studies were used to investigate the impact of the synaptic background activity observed in neocortical neurons in vivo. We simulated background activity in vitro using two stochastic Ornstein-Uhlenbeck processes describing glutamatergic and GABAergic synaptic conductances, which were injected into a cell in real time using the dynamic clamp technique. With parameters chosen to mimic in vivo conditions, layer 5 rat prefrontal cortex cells recorded in vitro were depolarized by about 15 mV, their membrane fluctuated with a S.D. of about 4 mV, their input resistances decreased five-fold, their spontaneous firing had a high coefficient of variation and an average firing rate of about 5-10 Hz. Brief changes in the variance of the alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) synaptic conductance fluctuations induced time-locked spiking without significantly changing the average membrane potential of the cell. These transients mimicked increases in the correlation of excitatory inputs. Background activity was highly effective in modulating the firing-rate/current curve of the cell: the variance of the simulated gamma-aminobutyric acid (GABA) and AMPA conductances individually set the input/output gain, the mean excitatory and inhibitory conductances set the working point, and the mean inhibitory conductance controlled the input resistance. An average ratio of inhibitory to excitatory mean conductances close to 4 was optimal in generating membrane potential fluctuations with high coefficients of variation. We conclude that background synaptic activity can dynamically modulate the input/output properties of individual neocortical neurons in vivo.

Fluctuating synaptic conductances recreate in vivo-like activity in neocortical neurons.

To investigate the basis of the fluctuating activity present in neocortical neurons in vivo, we have combined computational models with whole-cell recordings using the dynamic-clamp technique. A simplified 'point-conductance' model was used to represent the currents generated by thousands of stochastically releasing synapses. Synaptic activity was represented by two independent fast glutamatergic and GABAergic conductances described by stochastic random-walk processes. An advantage of this approach is that all the model parameters can be determined from voltage-clamp experiments. We show that the point-conductance model captures the amplitude and spectral characteristics of the synaptic conductances during background activity. To determine if it can recreate in vivo-like activity, we injected this point-conductance model into a single-compartment model, or in rat prefrontal cortical neurons in vitro using dynamic clamp. This procedure successfully recreated several properties of neurons intracellularly recorded in vivo, such as a depolarized membrane potential, the presence of high-amplitude membrane potential fluctuations, a low-input resistance and irregular spontaneous firing activity. In addition, the point-conductance model could simulate the enhancement of responsiveness due to background activity. We conclude that many of the characteristics of cortical neurons in vivo can be explained by fast glutamatergic and GABAergic conductances varying stochastically.

A model of the inward current Ih and its possible role in thalamocortical oscillations.

We investigated the kinetic properties of the hyperpolarization-activated inward current (Ih) of thalamocortical (TC) neurons. Recently, it was shown that this current is characterized by different time constants of activation and inactivation, which was in apparent conflict with the single-exponential time course of the current. We introduce here a model of Ih based on the cooperation of a slow and a fast activation variable and show that this kinetic scheme accounts for these apparently conflicting experimental data. We also report that following the combination of such a current with other currents seen in TC cells, one observes several types of oscillating behavior, similar to the slow oscillations and the spindle-like oscillations seen in vitro.

Modeling the control of reticular thalamic oscillations by neuromodulators.

Compartmental models of thalamic reticular (RE) neurons were investigated based on current-clamp and voltage-clamp data. Spontaneous oscillations in the model arise from the interaction between inhibitory synaptic currents and the rebound burst of RE cells. These oscillations critically depend on the level of the resting membrane potential. A network of RE neurons can be switched between silent and sustained oscillatory behavior by modulating a leak potassium current through neuromodulatory synapses. These results suggest that neuromodulators, such as noradrenaline, serotonin and glutamate, can exert a decisive control over the oscillatory activity of systems of RE cells. The model may explain why the isolated RE nucleus oscillates spontaneously in vivo but not in vitro.

Pacemaker-Induced Coherence in Cortical Networks.

A simple mathematical model of cortical tissue is introduced and the system's dynamics is monitored when a small subset of neurons is submitted to oscillatory inputs of various frequency and waveform. In the absence of input, the system shows desynchronized or "turbulent" behavior. The oscillatory input synchronizes the neuronal activity, which is strongest for inputs of low frequency. The increase of spatial coherence is estimated from the spatial autocorrelation function whereas the increase in temporal coherence is evaluated from correlation dimensions. The model accounts qualitatively for some of the features of the thalamocortical system.

Modelling corticothalamic feedback and the gating of the thalamus by the cerebral cortex.

Morphological studies have shown that excitatory synapses from the cortex constitute the major source of synapses in the thalamus. However, the effect of these corticothalamic synapses on the function of the thalamus is not well understood because thalamic neurones have complex intrinsic firing properties and interact through multiple types of synaptic receptors. Here we investigate these complex interactions using computational models. We show first, using models of reconstructed thalamic relay neurones, that the effect of corticothalamic synapses on relay cells can be similar to that of afferent synapses, in amplitude, kinetics and timing, although these synapses are located in different regions of the dendrites. This suggests that cortical EPSPs may complement (or predict) the afferent information. Second, using models of reconstructed thalamic reticular neurones, we show that high densities of the low-threshold Ca2+ current in dendrites can give these cells an exquisite sensitivity to cortical EPSPs, but only if their dendrites are hyperpolarized. This property has consequences at the level of thalamic circuits, where corticothalamic EPSPs evoke bursts in reticular neurones and recruit relay cells predominantly through feedforward inhibition. On the other hand, with depolarized dendrites, thalamic reticular neurones do not generate bursts and the cortical influence on relay cells is mostly excitatory. Models therefore suggest that the cortical influence can either promote or antagonize the relay of information, depending on the state of the dendrites of reticular neurones. The control of these dendrites may therefore be a determinant of attentional mechanisms. We also review the effect of corticothalamic feedback at the network level, and show how the cortical control over the thalamus is essential in co-ordinating widespread, coherent oscillations. We suggest mechanisms by which different modes of corticothalamic interaction would allow oscillations of very different spatiotemporal coherence to coexist in the thalamocortical system.

Why do we sleep?

Slow-wave sleep consists in slowly recurring waves that are associated with a large-scale spatio-temporal synchrony across neocortex. These slow-wave complexes alternate with brief episodes of fast oscillations, similar to the sustained fast oscillations that occur during the wake state. We propose that alternating fast and slow waves consolidate information acquired previously during wakefulness. Slow-wave sleep would thus begin with spindle oscillations that open molecular gates to plasticity, then proceed by iteratively 'recalling' and 'storing' information primed in neural assemblies. This scenario provides a biophysical mechanism consistent with the growing evidence that sleep serves to consolidate memories.