Wake-like activities and electrical silences in the human sleeping brain: functional roles and spatio-temporal dynamics in the thalamo-cortical network

As pointed out by several papers, the less than one hertz oscillations or Sleep Slow Oscillations (SSOs) are the electrophysiological stigmata of the mammalian sleep. This cellular behavior, mainly involving thalamus and cortex, consisted of hyperpolarized phases (lasting 500 msec, down states) followed by depolarized ones (lasting 500 msec, up states). The electrical silence during down states, on the one hand, prevents any synaptic and network activity and on the other hand, creates the ionic conditions for a rebound of neural discharge (huge synaptic and network activity during up state). The presence of down states clearly marks the phenomenon of cortical bistability, which in turn reflects a deep hyperpolarization sustained by the opening of different K+-channels. According to the Integrated Information Theory of Giulio Tononi, down states prevent the emergence of largescale neural integrations and thus induce the break down of functional connectivity. This allows a functional segregation of independent cortical modules, which represents the condicio sine qua non for sleep unconsciousness. Other functional roles endowed in the SSO are memory consolidation and synaptic downscaling. The aim of this thesis is to investigate, via EEG, in human spontaneous and evoked SSOs: (i) the relationships between wake-like activities and electrical silence; (ii) the role of the thalamus; (iii) the quenching of sensory processing and thus of consciousness. Regarding point (i) we have found a positive bump preceding the down state characterized by an increase of high frequency activities. The presence of this high frequency activity before down state suggests a cortical ignition mechanism for the spontaneous SSO. As far as point (ii) is concerned, we have investigated how the thalamus influences the cortical expression of the SSOs. To this aim, we have studied SSO features in a case of Fatal Familial Insomnia (FFI) with a selective thalamic neurodegeneration of nuclei mainly involved in spindle generation. In the FFI patient, we have found a reduction of SSO event rate, some morphological alterations of SSO structure, and a significant reduction in grouping high frequency activity during up state. As for point (iii), we studied K-Complexes (KCs), namely SSOs evoked by sensory stimulations. The main results of this study are: a positive wave (P200) precedes the down state (N550); the topology of P200 latency depends on the sensory modality of stimulation (acoustic, tactile and visual) with earliest waves in the related primary sensory areas; the P200 travels as a cortical excitation inducing N550 and P900 (up state) in associative fronto-central areas; when KCs are not evoked the P200-like excitations have lower amplitude compared to evoked KC P200; the down state latency topology is affected by the proneness to bistability, i.e. the amount of K+-channel that favor a synchronized falling into down state. As a whole the results of the thesis indicate that Slow Wave Sleep (SWS) is not a mere quiescent state but rather an active state in which changes of neural dynamics allow a well orchestrated interplay of unconscious behavior and memory consolidation. The final consequence is the maintenance of homeostasis. The SSO is the cellular phenomenon capable to coalesce wake-like activities and electrical silences, synthesizing at microscopic level the macroscopic complexity of SWS. This thesis allowed exploring thalamo-cortical dynamics by studying spontaneous and evoked SSOs. In synthesis the human-environment interaction (including visceral stimuli) during sleep overlaps that of wakefulness, since thalamus and cortical areas devoted to the first step of sensory processing are identical. The difference between wake and sleep is only sustained by the down state. In conclusion the study of SSO clarifies many issues linked to sleep and in particular to the real efficacy of a good sleep. This opens the door to the application of SSO study in different preclinical or clinical conditions.