We describe a novel slow oscillation in intracellular recordings from cortical association areas 5 and 7, motor areas 4 and 6, and visual areas 17 and 18 of cats under various anesthetics. The recorded neurons (n = 254) were antidromically and orthodromically identified as corticothalamic or callosal elements receiving projections from appropriate thalamic nuclei as well as from homotopic foci in the contralateral cortex. Two major types of cells were recorded: regular-spiking (mainly slow-adapting, but also fast-adapting) neurons and intrinsically bursting cells. A group of slowly oscillating neurons (n = 21) were intracellularly stained and found to be pyramidal-shaped cells in layers III-VI, with luxuriant basal dendritic arbors. The slow rhythm appeared in 88% of recorded neurons. It consisted of slow depolarizing envelopes (lasting for 0.8-1.5 sec) with superimposed full action potentials or presumed dendritic spikes, followed by long-lasting hyperpolarizations. Such sequences recurred rhythmically at less than 1 Hz, with a prevailing oscillation between 0.3 and 0.4 Hz in 67% of urethane-anesthetized animals. While in most neurons (approximately 70%) the repetitive spikes superimposed on the slow depolarization were completely blocked by slight DC hyperpolarization, 30% of cells were found to display relatively small (3-12 mV), rapid, all-or-none potentials after obliteration of full action potentials. These fast spikes were suppressed in an all-or-none fashion at Vm more negative than -90 mV. The depolarizing envelope of the slow rhythm was reduced or suppressed at a Vm of -90 to -100 mV and its duration was greatly reduced by administration of the NMDA blocker ketamine. In keeping with this action, most (56%) neurons recorded in animals under ketamine and nitrous oxide or ketamine and xylazine anesthesia displayed the slow oscillation at higher frequencies (0.6-1 Hz) than under urethane anesthesia (0.3-0.4 Hz). In 18% of the oscillating cells, the slow rhythm mainly consisted of repetitive (15-30 Hz), relatively short-lasting (15-25 msec) IPSPs that could be revealed by bringing the Vm at more positive values than -70 mV. The long-lasting (approximately 1 sec) hyperpolarizing phase of the slow oscillation was best observed at the resting Vm and was reduced at about -100 mV. Simultaneous recording of another cell across the membrane demonstrated synchronous inhibitory periods in both neurons. Intracellular diffusion of Cl- or Cs+ reduced the amplitude and/or duration of cyclic long-lasting hyperpolaryzations.(ABSTRACT TRUNCATED AT 400 WORDS)

The newly described slow cortical rhythm (approximately 0.3 Hz), whose depolarizing-hyperpolarizing components are analyzed in the preceding article, is now investigated from the standpoint of its relations with delta (1-4 Hz) and spindle (7-14 Hz) rhythmicity. Regular-spiking and intrinsically bursting cortical neurons were mostly recorded from association suprasylvian areas 5 and 7; fewer neurons were also recorded from pericruciate motor and posterolateral visual areas. Although most cells were investigated under various anesthetics, a similar slow cortical rhythm was found in animals with brainstem transection at the low- or high-collicular levels. These cerveau isolé (isolated forebrain) preparations display the major sleep rhythms of the EEG in the absence of general anesthetics. In 38% of recorded cortical neurons (n = 105), the slow rhythm was combined with delta oscillation. Both cellular rhythms were phase locked to the slow and delta oscillations in the surface- and depth-recorded EEG. In a group of this cell sample (n = 47), delta activity occurred as stereotyped, clock-like action potentials during the interdepolarization lulls of the slow rhythm. In another neuronal subsample (n = 58), delta events were grouped in sequences superimposed upon the depolarizing envelope of the slow rhythm, with such sequences recurring rhythmically at approximately 0.3-0.4 Hz. The associations between the two cellular and EEG rhythms (1-4 Hz and 0.3-0.4 Hz) were quantified by means of autocorrelograms, cross-correlograms, and spike-triggered averages. In 26% of recorded neurons (n = 72), the slow rhythm was combined with spindle oscillations. Regular-spiking cortical neurons fully reflected the whole frequency range of thalamically generated spindles (7-14 Hz). However, during similar patterns of EEG spindling, intrinsically bursting cells fired grouped action potentials (with intraburst frequencies of 100-200 Hz) at only 2-4 Hz. The dependence of the slow cortical oscillation upon the thalamus was studied by lesions and stimulation. The slow rhythm survived extensive ipsilateral thalamic destruction by means of electrolytic lesions or kainate-induced loss of perikarya in thalamic nuclei that were input sources to the recorded cortical neurons. To further prevent the possibility of a thalamic role in the genesis of the slow rhythm, through the contralateral thalamocortical systems and callosal projections, we also transected the corpus callosum in thalamically lesioned animals, and still recorded the slow rhythm in cortical neurons. These data indicate that the thalamus is not essentially implicated in the genesis of the slow rhythm.(ABSTRACT TRUNCATED AT 400 WORDS)

We investigated the synchronization of fast spontaneous oscillations (mainly 30-40 Hz) in anesthetized and behaving cats by means of simultaneous extra- and intracellular recordings from multiple neocortical areas. Fast Fourier transforms, auto- and cross-correlations, and spike- or wave-triggered averages were used to determine the frequency and temporal coherence of fast oscillations that outlasted the stimulation of ascending activating systems or that occurred naturally during behavioral states of waking and rapid eye movement (REM) sleep but also appeared during the depolarizing phases of slow sleep oscillations. In 90% of microelectrode tracks, the fast oscillations did not show field reversal at any depth of the cortex and were not observable in the underlying white matter. The negative field potentials of the fast oscillations were associated at all depths with neuronal firing. This field potential property of fast oscillations was in sharp contrast to the reversal of slow sleep oscillation or evoked potentials at depths of 0.25-0.5 mm. The coherence of fast spontaneous rhythms was spatially limited, being confined within a cortical column and among closely located neocortical sites, in contrast to the long-range synchronization of slow sleep rhythms. Depolarizing current pulses elicited spike-bursts (200-400 Hz) recurring at a frequency of 30-40 Hz. Our experiments demonstrate that the conventional notion of a totally desynchronized cortical activity upon arousal should be revised as fast rhythms are enhanced and synchronized within intracortical networks during brain activation. Spontaneously occurring, subthreshold membrane potential depolarizing oscillations may bias cortical and thalamic neurons to respond synchronously, at fast frequencies, to relevant stimuli in the wake state or to internally generated drives in REM sleep.

No abstract

The synchronization of fast (mainly 30 to 40 Hz) oscillations in intrathalamic and thalamocortical (TC) networks of cat was studied under ketamine-xylazine anesthesia and in behaving animals by means of field potential, extra- and intracellular recordings from multiple sites in the thalamic reticular (RE) nucleus, dorsal (sensory, motor, and intralaminar) thalamic nuclei, and related neocortical areas. Far from being restricted to tonically activated behavioral states, the fast oscillations also appeared during resting sleep and deep anesthesia, when they occurred over the depolarizing component of the slow (<1 Hz) oscillation and were suppressed during the prolonged hyper-polarizations of RE, TC, and cortical neurons. The synchronization of fast rhythms among different thalamic foci was robust. Fast rhythmic cortical waves and subthreshold depolarizing potentials in TC neurons were highly coherent; however, the synchronization of the fast oscillation required recordings from reciprocally related neocortical and thalamic foci, as identified by monosynaptic responses in both directions. The short-range spatial confinement of coherent fast rhythms contrasted with the large-scale synchronization of low-frequency sleep rhythms. Transient fast rhythms, appearing over the depolarizing envelope of the slow sleep oscillation, became sustained when brain activation was elicited by stimulation of mesopontine cholinergic nuclei or during brain-active behavioral states in chronic experiments. These data demonstrate that fast rhythms are part of the background electrical activity of the brain and that desynchronization, used to designate brain-active states, is an erroneous term inasmuch as the fast oscillations are synchronized not only in intracortical but also in intrathalamic and TC networks.