Search

Sleep spindles are generated in thalamocortical circuits. Thalamocortical projections can be classified as one of two main types: a “core” network consisting of spatially concentrated projections terminating in cortical layer IV, and a “matrix” network with more diffuse projections, terminating in the superficial cortical layers. It has been hypothesized that sleep spindles can be generated in both systems, but localized spindles are preferentially generated in the “core” system and global spindles in the “matrix system”. Our analysis of laminar microelectrode recordings during ECoG spindles indeed reveals that current source density (CSD) is maximal in layer I-II and IV, consistent with the anatomy of the hypothesized “core” and “matrix” networks. However, CSD is prominent at both depths in both local and global spindles, and CSD magnitude is strongly correlated between superficial layers and layer IV (although less so in local spindles). Fast spindles slow relatively greater CSD in layer IV and slow spindles in superficial layers. Sleep spindles are associated with several neurobehavioral markers, such as memory consolidation, cognitive ability and disease phenotypes. These associations, however, were mainly found with fast spindles. Centroparietal fast spindles also appear more prominently in EEG recordings and leave more prominent spectral peaks. While neither slow nor fast spindles rely exclusively on both “core” and “matrix” thalamocortical projections, it is likely that spatially focused “core” projections to layer IV are especially significant for sleep spindle function due to their role in fast spindle generation.

NREM sleep slow waves reflect both the temporal dynamics of the synchronized onset and cessation of cortical neuron firing and the spatial extent of this synchronized firing pattern. This synchronization is crucially dependent on synaptic strength, and the renormalization of synaptic changes occurring in wakefulness is thought to be one of the main functions of NREM sleep. In line with this hypothesis, slow waves are more frequent, larger and steeper in cases when the synaptic load is high and the need for homeostatic synaptic renormalization is strong: early in the night, in younger subjects with greater neural plasticity and after prolonged or more intensive wakefulness. Slow waves exhibit substantial heterogeneity and may contribute to these processes to a differing degree. In NREM sleep, series of large-amplitude slow waves (Cyclic Alternating Pattern [CAP] A1) sometimes appear, either spontaneously with predictably cyclicity or after sensory stimulation. Slow waves in CAP A1 resemble slow waves normally recorded under high homeostatic sleep pressure: they are characterized by a greater spatial extent, larger amplitude, steeper slopes (even when correcting for amplitude) and higher EEG synchronization, even though their topography and propagation patterns are similar to ordinary (non-CAP) slow waves. These results indicate that CAP A1 represents a period of highly synchronous neuronal firing over large areas of the cortical mantle. This feature may contribute to the role CAP A1 plays in both normal synaptic homeostasis and – through greater synchronization – in the generation of epileptiform phenomena in epileptic patients.

Introduction Sleep spindles are generated in thalamo-cortical networks. Thalamo-cortical projections include core projections to middle and matrix projections to superficial cortical layers. It has been proposed that sleep spindles can be generated in either of these networks. Materials and methods We used electrophysiological data recorded with subdural grid and intracortical multielectrodes in five therapy-resistant epileptic patients undergoing invasive monitoring. Spindles were detected automatically in electrocorticography channels and detection-triggered laminar microelectrode recordings were analyzed. The second spatial derivative of spindle-band filtered (band pass 10–16 Hz, 96 dB) local field potential data allowed to approximate the laminar distribution of the transmembrane generators (current source density, CSD) of spindles. The information regarding the concurrent population cellular activity was obtained by applying a high frequency band filter (500–5000 Hz, 48 dB) on epoched data. Based on the histological reconstructions of laminar electrode tracks, we could identify the precise laminar localization of current sources and sinks. Results Results indicate highly heterogeneous intracortical local field potentials during sleep spindles, and maximum CSD in layer I, II and IV. Multi-unit activity was increased in spindle troughs as it was identified based on the spindle band filtered local field potentials. The single unit analyis revealed preferential firing of most neurons during the positive-negative phase transition of spindles. Conclusions The results confirm that both core and matrix thalamocortical projections contribute to spindle generation, but most spindles appear to be generated by a combination of the two and only a small minority by only one or the other. Neuronal firing tracks the phase of spindles in corticographic channels. Supported by KTIA_NAP_13-1-2013-0001.

Introduction Deep brain stimulation of anterior nucleus of the thalamus (ANT-DBS) is an approved therapeutic approach for drug resistant epilepsy. Little is known about the participation of ANT in epileptic and sleep EEG processes. The appearance of the sleep spindles (SS) is related to interictal discharges (IID), and thalamus is considered as the generator of SSs, however, according to animal data, ANT do not participate in the genesis of sleep spindles. Methods In this present study we included 10 ANT-DBS bitemporal epilepsy patients with externalised leads and surface electrodes undergoing 2 days video-EEG monitoring (VEM). We investigated IIDs, ictal events (IE) and SSs in ANT and also in surface EEG recordings. IIDs and IEs were analysed by visual inspection, SSs were detected according to the patient own special spindle frequency, separately on the scalp (Fp1-2, F3-4, F7-8) and thalamus both sides. SS spectral peak frequency and synchronization between scalp and ANT was measured. Results IIDs appeared in the thalamus and on the scalp both dependent and independent manner. Three patients had seizures during VEM. Seizures did not start but start spread to the ANT electrodes. The slow, frontal type of SSs were prevailing in the ANT. Conclusion ANT participates in both sleep and epileptogenic processes. It seems that IIDs and IE spread to the ANT. Frontal, slow frequency SSs has local sources int he ANT indication its participation in sleep processes. Further evaluation for seizure outcome of ANT-DBS therapy and correlation with ANT recordings are underway.