The slow (< 1 Hz) oscillation in reticular thalamic and thalamocortical neurons: scenario of sleep rhythm generation in interacting thalamic and neocortical networks

M Steriade, D Contreras, RC Dossi… - Journal of …, 1993 - Soc Neuroscience
M Steriade, D Contreras, RC Dossi, A Nunez
Journal of Neuroscience, 1993Soc Neuroscience
As most afferent axons to the thalamus originate in the cerebral cortex, we assumed that the
slow (< 1 Hz) cortical oscillation described in the two companion articles is reflected in
reticular (RE) thalamic and thalamocortical cells. We hypothesized that the cortically
generated slow rhythm would appear in the thalamus in conjunction with delta and spindle
oscillations arising from intrinsic and network properties of thalamic neurons. Intracellular
recordings have been obtained in anesthetized cats from RE (n= 51) and cortically …
As most afferent axons to the thalamus originate in the cerebral cortex, we assumed that the slow (< 1 Hz) cortical oscillation described in the two companion articles is reflected in reticular (RE) thalamic and thalamocortical cells. We hypothesized that the cortically generated slow rhythm would appear in the thalamus in conjunction with delta and spindle oscillations arising from intrinsic and network properties of thalamic neurons. Intracellular recordings have been obtained in anesthetized cats from RE (n = 51) and cortically projecting (n = 240) thalamic neurons. RE cells were physiologically identified by cortically evoked high-frequency spike bursts and depolarizing spindle oscillations. Thalamocortical cells were recognized by backfiring from appropriate neocortical areas, spindle- related cyclic IPSPs, and hyperpolarization-activated delta oscillation consisting of rhythmic low-threshold spikes (LTSs) alternating with afterhyperpolarizing potentials (AHPs). The slow rhythm (0.3–0.5 Hz) was recorded in 65% of RE neurons. In approximately 90% of oscillating cells, the rhythm consisted of prolonged depolarizations giving rise to trains of single action potentials. DC hyperpolarization increased the synaptic noise and, in a few cells, suppressed the long-lasting depolarizing phase of the slow rhythm, without blocking the fast EPSPs. In approximately 10% of oscillating neurons, the hyperpolarizing phase of the oscillation was much more pronounced, thus suggesting that the slow rhythm was produced by inhibitory sculpturing of the background firing. The slow oscillation was associated with faster rhythms (4–8 Hz) in the same RE neuron. The slow rhythm of RE neurons was closely related to EEG wave complexes recurring with the same frequency, and its strong dependency upon a synchronized state of cortical EEG was observed during shifts in EEG patterns at different levels of anesthesia. In 44% of thalamocortical cells the slow rhythm of depolarizing sequences was apparent and it could coexist with delta or spindle oscillations in the same neuron. The occurrence of the slowly recurring depolarizing envelopes was delayed by the hyperpolarizing spindle sequences or by the LTS-AHP sequences of delta oscillation. The hyperpolarization-activated delta potentials that tended to dampen after a few cycles were grouped in sequences recurring with the slow rhythm. We finally propose a unified scenario of the genesis of the three major sleep rhythms: slow, delta, and spindle oscillations.
Soc Neuroscience