Low-frequency rhythms in the thalamus of intact-cortex and decorticated cats

1. The patterns and synchronization of low-frequency, sleeplike rhythms (slow, spindle and delta oscillations) were compared in the intact-cortex and decorticated hemispheres of cats under ketamine-xylazine anesthesia. Intracellular recordings were performed in intact and decorticated hemispheres from 58 rostrolateral thalamic reticular (RE) neurons and from 164 thalamocortical (TC) neurons in the ventrolateral (VL) nucleus. In the decorticated hemisphere, dual intracellular recordings were performed from five RE-VL cell couples and from 12 TC cell couples within the VL nucleus. In addition, field potentials were simultaneously recorded from the neocortex (electroencephalogram) and ipsilateral thalamus [electrothalamogram (EThG)] of the intact (right) hemisphere, while EThG was recorded from the VL nucleus of the decorticated (left) hemisphere. 2. The slow oscillation (< 1 Hz) was absent in all 72 VL cells and in 23 of 25 RE cells from the decorticated hemisphere, as well as in the EThG recorded from the VL nucleus in the decorticated hemisphere, whereas it was simultaneously present in the cortex and thalamus of the intact hemisphere. The remaining two RE neurons (8%) in the decorticated hemisphere oscillated in close time relation with the slow oscillation in the cortex and thalamus of the opposite hemisphere; averaged activities showed that the onset of depolarization in RE cell followed 12 ms after the sharp depth-negative (depolarizing) component in the contralateral cortex. We view this result as the electrophysiological correlate of a disynaptic excitatory pathway consisting of crossed cortical projections, first relayed in contralateral dorsal thalamic nuclei. 3. The patterns of thalamic spindles (7–14 Hz) differed between the two hemispheres. Whereas the decorticated hemisphere displayed prolonged, waxing and waning spindles, the spindles in the intact-cortex hemisphere were short and exclusively waning and followed the depth-negative component of cortical slow oscillation. This result indicates that the synchronized corticothalamic drive associated with the slow oscillation fully entrains thalamic circuits from the onset of spindles, thus preventing further waxing. Similar differences between waxing and waning and waning spindles were obtained by stimulating with different intensities the thalamus in the decorticated hemisphere. 4. Simultaneous intracellular recordings from two VL cells or from RE and VL cells showed nearly simultaneous spindle sequences in the decorticated hemisphere. 5. The hyperpolarization-activated intrinsic delta oscillation (1–4 Hz) of TC cells was asynchronous in the decorticated hemisphere. 6. These results strengthen the idea that the slow oscillation is cortical in origin; demonstrate a full, short-range, intrathalamic synchrony of spindles in the absence of cortex; and indicate that the pattern of spindles, a sleep rhythm that is conventionally regarded as purely thalamic, is shaped by the corticothalamic feedback.