Journal of Neuroscience, Vol 9, 2215-2229, Copyright © 1989 by Society for Neuroscience
Phasic activation of lateral geniculate and perigeniculate thalamic neurons during sleep with ponto-geniculo-occipital waves
M Steriade, D Pare, D Bouhassira, M Deschenes and G Oakson
Laboratoire de Neurophysiologie, Faculte de Medecine, Universite Laval, Quebec, Canada.
Ponto-geniculo-occipital (PGO) waves are spiky field potentials generated
in cholinergic nuclei of the dorsolateral mesopontine tegmentum just prior
to and during rapid-eye-movement (REM) sleep and transferred toward
thalamic nuclei. These events are commonly regarded as physiological
correlates of oneiric behavior. We have examined the PGO-related discharges
of physiologically identified neurons located in the dorsal lateral
geniculate (LG) nucleus and perigeniculate (PG) sector of the reticular
thalamic complex in chronically implanted, naturally sleeping cats. PGO
focal waves and associated unit discharges were simultaneously recorded by
the same microelectrode. PGO waves herald the other signs of REM sleep (EEG
desynchronization and muscular atonia), appearing 30-90 sec before REM
sleep over the EEG-synchronized activity of slow-wave sleep (pre-REM
epoch). (1) Most PG neurons discharged bursts of action potentials in
relation to PGO waves during both pre-REM and REM sleep. (2) The
PGO-related activity of LG neurons was quite different. During the pre-REM
stage, PGO waves correlated with a short (7-15 msec), high-frequency
(300-500 Hz) spike burst of LG neurons, followed by a long (0.2-0.4 sec)
train of single spikes, whereas during REM sleep, the PGO-related activity
lacked the initial burst and consisted of a spike train that only slightly
exceeded the tonically increased background firing of LG cells. The
stereotyped characteristics of the PGO-related spike bursts during the
pre-REM epoch suggest that they are the extracellular reflection of a low-
threshold spike deinactivated by the tonic membrane hyperpolarization of LG
cells associated with the EEG-synchronized sleep state. Such bursts are
inactivated during the tonic depolarization of LG cells that occurs in REM
sleep. The synchronous spike bursts discharged by LG cells in relation with
the PGO waves of the pre-REM epoch probably underlie the much larger
amplitude of the PGO waves of the pre-REM epoch as compared with those of
the REM-sleep state. Since LG neurons have relatively low spontaneous
firing rates during the EEG- synchronized pre-REM epoch, the PGO-related
activity of this transitional stage leads to a higher signal-to-noise ratio
in the visual thalamocortical channel than during REM sleep. We suggest
that the PGO-related activity during the pre-REM epoch is related to vivid
imagery during this stage of sleep.