Interactions between hypocretinergic and GABAergic systems in the control of activity of neurons in the cat pontine reticular formation.

Anatomical studies have demonstrated that hypocretinergic and GABAergic neurons innervate cells in the nucleus pontis oralis (NPO), a nucleus responsible for the generation of active (rapid eye movement (REM)) sleep (AS) and wakefulness (W). Behavioral and electrophysiological studies have shown that hypocretinergic and GABAergic processes in the NPO are involved in the generation of AS as well as W. An increase in hypocretin in the NPO is associated with both AS and W, whereas GABA levels in the NPO are elevated during W. We therefore examined the manner in which GABA modulates NPO neuronal responses to hypocretin. We hypothesized that interactions between the hypocretinergic and GABAergic systems in the NPO play an important role in determining the occurrence of AS or W. To determine the veracity of this hypothesis, we examined the effects of the juxtacellular application of hypocretin-1 and GABA on the activity of NPO neurons, which were recorded intracellularly, in chloralose-anesthetized cats. The juxtacellular application of hypocretin-1 significantly increased the mean amplitude of spontaneous EPSPs and the frequency of discharge of NPO neurons; in contrast, the juxtacellular microinjection of GABA produced the opposite effects, i.e., there was a significant reduction in the mean amplitude of spontaneous EPSPs and a decrease in the discharge of these cells. When hypocretin-1 and GABA were applied simultaneously, the inhibitory effect of GABA on the activity of NPO neurons was reduced or completely blocked. In addition, hypocretin-1 also blocked GABAergic inhibition of EPSPs evoked by stimulation of the laterodorsal tegmental nucleus. These data indicate that hypocretin and GABA function within the context of a neuronal gate that controls the activity of AS-on neurons. Therefore, we suggest that the occurrence of either AS or W depends upon interactions between hypocretinergic and GABAergic processes as well as inputs from other sites that project to AS-on neurons in the NPO.

Excitatory projections from the amygdala to neurons in the nucleus pontis oralis in the rat: an intracellular study.

There is a consensus that active (REM) sleep (AS) is controlled by cholinergic projections from the laterodorsal and pedunculopontine tegmental nuclei (LDT/PPT) to neurons in the nucleus pontis oralis (NPO) that generate AS (i.e. AS-Generator neurons). The present study was designed to provide evidence that other projections to the NPO, such as those from the amygdala, are also capable of inducing AS. Accordingly, the responses of neurons, recorded intracellularly in the NPO, were examined following stimulation of the ipsilateral central nucleus of the amygdala (CNA) in urethane-anesthetized rats. Single pulse stimulation in the CNA produced an early, fast depolarizing potential (EPSP) in neurons within the NPO. The mean latency to the onset of these excitatory postsynaptic potentials (EPSPs) was 3.6±0.2 ms. A late, small-amplitude inhibitory synaptic potential (IPSP) was present following EPSPs in a portion of the NPO neurons. Following stimulation of the CNA with a train of 8-10 pulses, NPO neurons exhibited a sustained depolarization (5-10 mV) of their resting membrane potential. When single subthreshold intracellular depolarizing current pulses were delivered to NPO neurons, CNA-induced EPSPs were sufficient to promote the discharge of these cells. Stimulation of the CNA with a short train of stimuli induced potent temporal facilitation of EPSPs in NPO neurons. Two forms of synaptic plasticity were revealed by the patterns of response of NPO neurons following stimulation of the CNA: paired-pulse facilitation (PPF) and post-tetanic potentiation (PTP). Six of recorded NPO neurons were identified morphologically with neurobiotin. They were medium to large, multipolar cells with diameters >20 µM, which resemble AS-on cells in the NPO. The present results demonstrate that amygdalar projections are capable of exerting a powerful excitatory postsynaptic drive that activates NPO neurons. Therefore, we suggest that the amygdala is capable of inducing AS via direct projections to AS-Generator neurons in the NPO.

A restricted parabrachial pontine region is active during non-rapid eye movement sleep.

The principal site that generates both rapid eye movement (REM) sleep and wakefulness is located in the mesopontine reticular formation, whereas non-rapid eye movement (NREM) sleep is primarily dependent upon the functioning of neurons that are located in the preoptic region of the hypothalamus. In the present study, we were interested in determining whether the occurrence of NREM might also depend on the activity of mesopontine structures, as has been shown for wakefulness and REM sleep. Adult cats were maintained in one of the following states: quiet wakefulness (QW), alert wakefulness (AW), NREM, or REM sleep induced by microinjections of carbachol into the nucleus pontis oralis (REM-carbachol). Subsequently, they were euthanized and single-labeling immunohistochemical studies were undertaken to determine state-dependent patterns of neuronal activity in the brainstem based upon the expression of the protein Fos. In addition, double-labeling immunohistochemical studies were carried out to detect neurons that expressed Fos as well as choline acetyltransferase, tyrosine hydroxylase, or GABA. During NREM, only a few Fos-immunoreactive cells were present in different regions of the brainstem; however, a discrete cluster of Fos+ neurons was observed in the caudolateral parabrachial region (CLPB). The number of Fos+ neurons in the CLPB during NREM was significantly greater (67.9±10.9, P<0.0001) compared with QW (8.0±6.7), AW (5.2±4.2), or REM-carbachol (8.0±4.7). In addition, there was a positive correlation (R=0.93) between the time the animals spent in NREM and the number of Fos+ neurons in the CLPB. Fos-immunoreactive neurons in the CLPB were neither cholinergic nor catecholaminergic; however, about 50% of these neurons were GABAergic. We conclude that a group of GABAergic and unidentified neurons in the CLPB are active during NREM and likely involved in the control of this behavioral state. These data open new avenues for the study of NREM, as well as for the explorations of interactions between these neurons that are activated during NREM and cells of the adjacent pontine tegmentum that are involved in the generation of REM sleep.

Nitrergic ventro-medial medullary neurons activated during cholinergically induced active (rapid eye movement) sleep in the cat.

The rostral ventro-medial medullary reticular formation is a complex structure that is involved with a variety of motor functions. It contains glycinergic neurons that are activated during active (rapid eye movement (REM)) sleep (AS); these neurons appear to be responsible for the postsynaptic inhibition of motoneurons that occurs during this state. We have reported that neurons in this same region contain nitric oxide (NO) synthase and that they innervate brainstem motor pools. In the present study we examined the c-fos expression of these neurons after carbachol-induced active sleep (C-AS). Three control and four experimental cats were employed to identify c-fos expressing nitrergic neurons using immunocytochemical techniques to detect the Fos protein together with neuronal nitric oxide synthase (nNOS) or nicotinamide adenine dinucleotide phosphate (NADPH)-diaphorase activity. The classical neurotransmitter content of the nitrergic cells in this region was examined through the combination of immunocytochemical techniques for the detection of glutamate, glycine, choline acetyltransferase (ChAT), tyrosine hydroxilase (TH) or GABA together with nNOS. During C-AS, there was a 1074% increase in the number of nitrergic neurons that expressed c-fos. These neurons did not contain glycine, ChAT, TH or GABA, but a subpopulation (15%) of them displayed glutamate-like immunoreactivity. Therefore, some of these neurons contain both an excitatory neurotransmitter (glutamate) and an excitatory neuromodulator (NO); the neurotransmitter content of the rest of them remains to be determined. Because some of the nitrergic neurons innervate brainstem motoneurons it is possible that they participate in the generation of tonic and excitatory phasic motor events that occur during AS. We also suggest that these nitrergic neurons may be involved in autonomic regulation during this state. In addition, because NO has trophic effects on target neurons, the present findings represent the first, albeit indirect, evidence for a possible trophic function of this nature during AS.

GABAergic processes in the mesencephalic tegmentum modulate the occurrence of active (rapid eye movement) sleep in guinea pigs.

The ventrolateral subdivision of the periaqueductal gray (vlPAG) and the adjacent dorsal mesencephalic reticular formation (dMRF) are involved in the modulation of active (rapid eye movement) sleep (AS). In order to determine the effects on AS of the suppression of neuronal activity in these regions, muscimol, a GABA receptor A (GABA(A)) receptor agonist, and bicuculline, a GABA(A) receptor antagonist, were microinjected bilaterally in guinea pigs and the states of sleep and wakefulness were examined. The main effect of muscimol was an increase in AS; this increase occurred in conjunction with a reduction in the time spent in wakefulness. The powerful effect of muscimol was striking especially when considering the small amount of naturally-occurring AS that is present in this species. Additional observable effects that were induced by muscimol were: 1) long lasting episodes of hypotonia/atonia during wakefulness and quiet sleep that included a lack of extensor tone in the hind limbs, and 2) frequently occurring cortical spindles, similar to those observed during naturally-occurring quiet sleep (sleep spindles), that were present during wakefulness. Conversely, bilateral microinjections of bicuculline induced a prolonged state of wakefulness and blocked the effect of subsequent injections of muscimol. These data suggest that endogenous GABA acts on GABA(A) receptors within the vlPAG and dMRF to promote AS in the guinea pig.

Brainstem glycinergic neurons and their activation during active (rapid eye movement) sleep in the cat.

It is well established that, during rapid eye movement (REM) sleep, somatic motoneurons are subjected to a barrage of inhibitory synaptic potentials that are mediated by glycine. However, the source of this inhibition, which is crucial for the maintenance and preservation of REM sleep, has not been identified. Consequently, the present study was undertaken to determine in cats the location of the glycinergic neurons, that are activated during active sleep, and are responsible for the postsynaptic inhibition of motoneurons that occurs during this state. For this purpose, a pharmacologically-induced state of active sleep (AS-carbachol) was employed. Antibodies against glycine-conjugated proteins were used to identify glycinergic neurons and immunocytochemical techniques to label the Fos protein were employed to identify activated neurons. Two distinct populations of glycinergic neurons that expressed c-fos were distinguished. One population was situated within the nucleus reticularis gigantocellularis (NRGc) and nucleus magnocellularis (Mc) in the rostro-ventral medulla; this group of neurons extended caudally to the ventral portion of the nucleus paramedianus reticularis (nPR). Forty percent of the glycinergic neurons in the NRGc and Mc and 25% in the nPR expressed c-fos during AS-carbachol. A second population was located in the caudal medulla adjacent to the nucleus ambiguus (nAmb), wherein 40% of the glycinergic cells expressed c-fos during AS-carbachol. Neither population of glycinergic cells expressed c-fos during quiet wakefulness or quiet (non-rapid eye movement) sleep. We suggest that the population of glycinergic neurons in the NRGc, Mc, and nPR participates in the inhibition of somatic brainstem motoneurons during active sleep. These neurons may also be responsible for the inhibition of sensory and other processes during this state. It is likely that the group of glycinergic neurons adjacent to the nucleus ambiguus (nAmb) is responsible for the active sleep-selective inhibition of motoneurons that innervate the muscles of the larynx and pharynx.

Neuronal mechanisms of active (rapid eye movement) sleep induced by microinjections of hypocretin into the nucleus pontis oralis of the cat.

Hypocretinergic (orexinergic) neurons in the hypothalamus project to the nucleus pontis oralis, a nucleus which plays a crucial role in the generation of active (rapid eye movement) sleep. We recently reported that the microinjection of hypocretin into the nucleus pontis oralis of chronically-instrumented, unanesthetized cats induces a behavioral state that is comparable to naturally-occurring active sleep. The present study examined the intracellular signaling pathways underlying the active sleep-inducing effects of hypocretin. Accordingly, hypocretin-1, a protein kinase C inhibitor and a protein kinase A inhibitor were injected into the nucleus pontis oralis in selected combinations in order to determine their effects on sleep and waking states of chronically instrumented, unanesthetized cats. Microinjections of hypocretin-1 into the nucleus pontis oralis elicited active sleep with a short latency. However, a pre-injection of bisindolylmaleimide-I, a protein kinase C-specific inhibitor, completely blocked the active sleep-inducing effects of hypocretin-1. The combined injection of bisindolylmaleimide-I and hypocretin-1 significantly increased the latency to active sleep induced by hypocretin-1; it also abolished the increase in the time spent in active sleep induced by hypocretin-1. On the other hand, the injection of 2'5'-dideoxyadenosine, an adenylyl cyclase inhibitor, did not block the occurrence of active sleep by hypocretin-1. We conclude that the active sleep-inducing effect of hypocretin in the nucleus pontis oralis is mediated by intracellular signaling pathways that act via G-protein stimulation of protein kinase C.

The role of tropomyosin-related kinase receptors in neurotrophin-induced rapid eye movement sleep in the cat.

The microinjection of nerve growth factor and neurotrophin-3 into the rostro-dorsal pontine tegmentum of the cat evokes a state that is comparable to naturally-occurring rapid eye movement sleep. Using two experimental paradigms, we tested the hypothesis that neurotrophin high-affinity receptors (trkA and trkC, tropomyosin-related kinase A and C, respectively) mediate this effect. First, trk and fos immunohistochemistry were combined to determine whether tyrosine kinase receptor-containing neurons in the dorsal pontine tegmentum are active in cats that exhibit long-lasting periods of rapid eye movement sleep following the local microinjection of nerve growth factor. During approximately two hours of recording, nerve growth factor-treated cats spent 59.8% of the time in a rapid eye movement sleep-like state; vehicle-injected (control) animals remained in quiet wakefulness and non-rapid eye movement sleep. Whereas control and nerve growth factor-treated cats exhibited a similar mean number of trkA- and trkC-immunoreactive neurons in the dorsal pontine tegmentum, the number of trkA- and trkC-immunoreactive neurons that expressed Fos, i.e. double-labeled cells that are presumably activated, was significantly larger in cats that were injected with nerve growth factor. Axon terminals contained tyrosine kinase receptor immunoreactivity in this region; many were apposed to Fos-immunoreactive neurons. In addition, patterns of tyrosine kinase receptor and Fos immunoreactivity similar to those observed in nerve growth factor-injected cats were present, in conjunction with long-lasting rapid eye movement sleep, following the microinjection of carbachol into the dorsal pons. In a second series of studies, nerve growth factor or neurotrophin-3 was injected alone or after K-252a, a blocker of tyrosine kinase receptors, into the rostro-dorsal pontine tegmentum. Nerve growth factor or neurotrophin-3 alone produced, with a mean latency of 4 min, a rapid eye movement sleep-like state. However, neurotrophin injections preceded by K-252a were not effective in inducing rapid eye movement sleep. These results indicate that the activation of trkA and trkC receptors in neurons in the pontine tegmentum is responsible, at least in part, for the rapid eye movement sleep-inducing effect of nerve growth factor and neurotrophin-3. Furthermore, the data suggest that these neurotrophins are capable of acting both pre- and postsynaptically to activate pontine neurons that are involved in the generation of rapid eye movement sleep.

Induction of rapid eye movement sleep by neurotrophin-3 and its co-localization with choline acetyltransferase in mesopontine neurons.

Because neurotrophin-3 (NT-3), a neurotrophic factor closely related to nerve growth factor, is capable of modulating neuronal activity [Yamuy et al., Neuroscience 95 (2000a) 1089-1100], we sought to examine if the microinjection of NT-3 into the nucleus reticularis pontis oralis (NPO) of chronically prepared cats also induced changes in behavior. In contrast to vehicle administration, NT-3 injection induced, with a mean latency of 4.7 min, long-duration episodes (mean, 21.6 min) of a state that was polygraphically indistinguishable from naturally occurring REM sleep. If NT-3 plays a physiologic role in the generation of REM sleep, then an endogenous source for this neurotrophin that is capable of controlling the activity of NPO neurons should exist. We therefore determined whether cholinergic neurons in the latero-dorsal and pedunculo-pontine tegmental (LDT and PPT) nuclei, which are involved in the initiation of REM sleep and project to the NPO, contained NT-3. Most, if not all, of the LDT-PPT cholinergic neurons exhibited NT-3 immunoreactivity. A portion (10%) of the NT-3+ neurons in the LDT-PPT were not cholinergic. The present data indicate that NT-3 rapidly modulates the activity of NPO neurons involved in REM sleep and that cholinergic neurons in the LDT and PPT contain NT-3. Taken together, these results support the hypothesis that NT-3 may be involved in the control of naturally occurring REM sleep.

The motor inhibitory system operating during active sleep is tonically suppressed by GABAergic mechanisms during other states.

The present study was undertaken to explore the neuronal mechanisms responsible for muscle atonia that occurs after the microinjection of bicuculline into the nucleus pontis oralis (NPO). Specifically, we wished to test the hypothesis that motoneurons are postsynaptically inhibited after the microinjection of bicuculline into the NPO and determine whether the inhibitory mechanisms are the same as those that are utilized during naturally occurring active (rapid eye movement) sleep. Accordingly, intracellular records were obtained from lumbar motoneurons in cats anesthetized with alpha-chloralose before and during bicuculline-induced motor inhibition. The microinjection of bicuculline into the NPO resulted in a sustained reduction in the amplitude of the spinal cord Ia-monosynaptic reflex. In addition, lumbar motoneurons exhibited significant changes in their electrophysiological properties [i.e., a decrease in input resistance and membrane time constant, a reduction in the amplitude of the action potential's afterhyperpolarization (AHP) and an increase in rheobase]. Discrete, large-amplitude inhibitory postsynaptic potentials (IPSPs) were also observed in high-gain recordings from lumbar motoneurons. These potentials were comparable to those that are only present during the state of naturally occurring active sleep. Furthermore, stimulation of the medullary nucleus reticularis gigantocellularis evoked a large-amplitude IPSP in lumbar motoneurons after, but never prior to, the injection of bicuculline; this reflects the pattern of motor responses that occur in conjunction with the phenomenon of "reticular response-reversal." The preceding changes in the electrophysiological properties of motoneurons, as well as the development of active sleep-specific IPSPs, indicate that lumbar motoneurons are postsynaptically inhibited following the intrapontine administration of bicuculline in a manner that is comparable to that which occurs spontaneously during the atonia of active sleep. The present results support the conclusion that the brain stem-spinal cord inhibitory system, which is responsible for motor inhibition during active sleep, can be activated by the injection of bicuculline into the NPO. These data suggest that the active sleep-dependent motor inhibitory system is under constant GABAergic inhibitory control, which is centered in the NPO. Thus during wakefulness and quiet sleep, the glycinergically mediated postsynaptic inhibition of motoneurons is prevented from occurring due to GABAergic mechanisms.

Importance of the environment in conditioned behavior.

The purpose of this study was to determine whether an entire experimental situation, and not an individual single stimulus, could be employed to generate a conditioned response. Experiments were conducted on four adult cats. Each cat was conditioned in two different experimental environments (Situation I and Situation II), which consisted of two compartments that differed with respect to color, shape and illumination. Experiments were carried out separately on each cat; experimental sessions, which lasted 30 min, were conducted two or three times a week. Two cats (Nos. 1 and 2) were first trained during 15 sessions in Situation I, and then during 15 subsequent sessions in Situation II. During each session in Situation I, electrical stimulation, which was applied to the basal forebrain area (BFA), evoked a slow-wave EEG pattern; in addition, the animals would close their eyes and lie down (i.e., they exhibited typical presleep behavior). After three to five sessions, this behavior began to appear as soon as the cats were introduced into the experimental compartment, even before stimulation was applied. In a further series of sessions, the cats were placed for 15 sessions in Situation II, wherein stimulation was applied to the lateral hypothalamus (LH). Stimulation of the LH evoked a desynchronized EEG pattern that was accompanied by excitatory behavior. In the other two cats (Nos. 3 and 4), the animals were first trained for 15 sessions in Situation II; subsequently, they were trained for 15 sessions in Situation I. Finally, a test of cross-stimulation was performed. Stimulation of the BFA (which was previously used in Situation I) was applied, one time only, to cats 1 and 2 in Situation II; stimulation of the LH (which was previously used in Situation II) was applied, once only, to cats 3 and 4 in Situation I. In both cases, the animals did not exhibit any of the previously observed behavioral reactions or EEG patterns of activity. The preceding results confirm our hypothesis that, in each situation, a conditioned reaction was established in response to the totality of the experimental environment.

Neuronal network analysis based on arrival times of active-sleep specific inhibitory postsynaptic potentials in spinal cord motoneurons of the cat.

The neuronal network responsible for motoneuron inhibition and loss of muscle tone during active (REM) sleep can be activated by the injection of the cholinergic agonist carbachol into a circumscribed region of the brainstem reticular formation. In the present report, we studied the arrival times of inhibitory postsynaptic potentials (IPSPs) observed in intracellular recordings from cat spinal cord motoneurons. These recordings were obtained during episodes of motor inhibition induced by carbachol or during motor inhibition associated with naturally occurring active sleep. When the observed IPSP arrival times were analyzed as a superposition of renewal processes occurring in a pool of pre-motor inhibitory interneurons, it was possible to estimate the following parameters: (1) the number of independent sources of the IPSPs; (2) the rate at which each source was bombarded with excitatory postsynaptic potentials (EPSPs); and (3) the number of EPSPs required to bring each source to threshold. From the data based upon the preceding parameters and the unusually large amplitudes of the active sleep-specific IPSPs, we suggest that each source is a cluster of synchronously discharging pre-motor inhibitory interneurons. The analysis of IPSP arrival times as a superposition of renewal processes, therefore, provides quantitative information regarding neuronal activity that is as far as two synapses upstream from the site of the recording electrode. Consequently, we suggest that a study of the temporal evolution of these parameters could provide a basis for dynamic analyses of this neuronal network and, in the future, for other neuronal networks as well.

Hypocretin (orexin) input to trigeminal and hypoglossal motoneurons in the cat: a double-labeling immunohistochemical study.

In trigeminal and hypoglossal motor nuclei of adult cats, hypocretin immunoreactive fiber varicosities were observed in apposition to retrogradely labeled motoneuron somata and dendrites. Among those lateral hypothalamus neurons that project to the hypoglossal nucleus some were determined to be hypocretin immunoreactive and were located amongst the single-labeled hypocretinergic neurons. These data suggest that hypocretin may play a role in the synaptic control of these motoneurons.

Effects on sleep and wakefulness of the injection of hypocretin-1 (orexin-A) into the laterodorsal tegmental nucleus of the cat.

Anatomical data demonstrate a dense projection, in the cat, from hypocretin (orexin) neurons in the hypothalamus to the laterodorsal tegmental nucleus (LDT), which is a critical pontine site that is involved in the regulation of the behavioral states of sleep and wakefulness. The present study was therefore undertaken to explore the hypocretinergic control of neurons in the LDT vis-à-vis these behavioral states. Accordingly, hypocretin-1 was microinjected into the LDT of chronic, unanesthetized cats and its effects on the percentage, latency, frequency and duration of wakefulness, quiet (non-REM) sleep and active (REM) sleep were determined. There was a significant increase in the time spent in wakefulness following the microinjection of hypocretin-1 into the LDT and a significant decrease in the time spent in active sleep. The increase in the percentage of wakefulness was due to an increase in the duration of episodes of wakefulness; the reduction in active sleep was due to a decrease in the frequency of active sleep episodes, but not in their duration. These data indicate that hypocretinergic processes in the LDT play an important role in both of the promotion of wakefulness and the suppression of active sleep.

Induction of wakefulness and inhibition of active (REM) sleep by GABAergic processes in the nucleus pontis oralis.

The present study was undertaken to explore the role of brainstem GABAergic processes in the control of the behavioral states of sleep and wakefulness, and to compare the effects of GABAA agonists and antagonists with those of GABAB agonists and antagonists on these behavioral states. Accordingly, the following drugs were microinjected into the nucleus pontis oralis (NPO) in chronic, unanesthetized cats: muscimol (GABAA agonist), bicuculline (GABAA antagonist), baclofen (GABAB agonist) and phaclofen (GABAB antagonist). The percentage, latency, frequency and duration of each behavioral state were measured in order to quantify the effects of these microinjections on wakefulness and sleep. Microinjections of either muscimol or baclofen immediately induced wakefulness. There was a significant increase in the duration and the percentage of time spent in wakefulness as well as an increase in the latency to active (REM) sleep. These changes were accompanied by a decrease in the percentage of time spent in active and quiet sleep. In contrast, injections of bicuculline or phaclofen produced active sleep. The percentage of time spent in active sleep and the frequency of active sleep increased while the percentage of time spent in wakefulness and the latency to active sleep was significantly reduced. The effects of GABAA receptor agonists and antagonists on wakefulness and active sleep were comparable, but stronger than those of GABAB receptor agonists and antagonists. These data indicate that pontine GABAergic processes acting on both GABAA and GABAB receptors play a critical role in generating and maintaining wakefulness and in controlling the occurrence of state of active sleep.

GABAergic neurons of the laterodorsal and pedunculopontine tegmental nuclei of the cat express c-fos during carbachol-induced active sleep.

The laterodorsal and pedunculopontine tegmental nuclei (LDT-PPT) are involved in the generation of active sleep (AS; also called REM or rapid eye movement sleep). Although the LDT-PPT are composed principally of cholinergic neurons that participate in the control of sleep and waking states, the function of the large number of GABAergic neurons that are also located in the LDT-PPT is unknown. Consequently, we sought to determine if these neurons are activated (as indicated by their c-fos expression) during active sleep induced by the microinjection of carbachol into the rostro-dorsal pons (AS-carbachol). Accordingly, immunocytochemical double-labeling techniques were used to identify GABA and Fos protein, as well as choline acetyltransferase (ChAT), in histological sections of the LDT-PPT. Compared to control awake cats, there was a larger number of GABAergic neurons that expressed c-fos during AS-carbachol (31.5+/-6.1 vs. 112+/-15.2, P<0.005). This increase in the number of GABA+Fos+ neurons occurred on the ipsilateral side relative to the injection site; there was a small decrease in GABA+Fos+ cells in the contralateral LDT-PPT. However, the LDT-PPT neurons that exhibited the largest increase in c-fos expression during AS-carbachol were neither GABA+ nor ChAT+ (47+/-22.5 vs. 228.7+/-14.0, P<0.0005). The number of cholinergic neurons that expressed c-fos during AS-carbachol was not significantly different compared to wakefulness. These data demonstrate that, during AS-carbachol, GABAergic as well as an unidentified population of neurons are activated in the LDT-PPT. We propose that these non-cholinergic LDT-PPT neurons may participate in the regulation of active sleep.

Orexin (hypocretin)-like immunoreactivity in the cat hypothalamus: a light and electron microscopic study.

Orexin-A-like immunoreactive (OrA-ir) neurons and terminals in the cat hypothalamus were examined using immunohistochemical techniques. OrA-ir neurons were found principally in the lateral hypothalamic area (LHA) at the level of the tuberal cinereum and in the dorsal and posterior hypothalamic areas. In the LHA the majority of the neurons were located dorsal and lateral to the fornix; a small number of OrA-ir neurons were also present in other regions of the hypothalamus. OrA-ir fibers with varicose terminals were detected in almost all hypothalamic regions. The high density of fibers was located in the suprachiasmatic nucleus, the infundibular nucleus (INF), the tuberomamillary nucleus (TM) and the supra- and pre-mamillary nuclei. Ultrastructural analysis revealed that OrA-ir neurons in the LHA receive abundant input from non-immunoreactive terminals. These terminals, which contained many small, clear, round vesicles with a few large, dense core vesicles, made asymmetrical synaptic contacts with OrA-ir dendrites, indicating that the activity of orexin neurons is under excitatory control. On the other hand, the terminals of OrA-ir neurons also made asymmetrical synaptic contact with dendrites in the LHA, the INF and the TM. The dendrites in the LHA were both non-immunoreactive and OrA-ir; conversely, the dendrites in the INF and the TM were non-immunoreactive. In these regions, OrA-ir terminals contained many small, clear, round vesicles with few large, dense core vesicles, suggesting that orexinergic neurons also provide excitatory input to other neurons in these regions.

Sleep and cognitive (memory) function: research and clinical perspectives.

The field of memory and sleep is controversial and extremely interesting, and the relationships between thought processes, i.e. cognition and sleep, have recently been examined in a variety of clinical and basic research settings, as well as being the object of intense interest by the general public. For example, there are data which demonstrate that insomnia, as well as specific sleep disorders, can have a negative impact on sleep cognition as well as affect daytime patterns of cognitive functioning. Thus, sleep, disturbed sleep and the lack of sleep appear to affect cognitive and memory functions. An International Workshop dealing with Sleep and Cognitive Function: Research and Clinical Perspectives was convened in Cancún, Mexico, 1-4 March 1999 under the auspices of the World Health Organization Worldwide Project on Sleep and Health and the World Federation of Sleep Research Societies. A great number of areas of intersection between sleep and cognitive function were examined during the course of the Workshop, such as aging, cognition and sleep and the dream process and sleep. The results of these discussions are included in a WHO publication (WHO Doc.: MSD/MBD/00.8). In the present report we concentrate on presenting a summary of a coherent set of data which examine memory consolidation during sleep and the impact of insomnia on cognitive functions. Based upon these data, a review of memory and drug effects that are sleep-related, and an examination of the relationship between hypnotics and cognitive function are included. Finally, a summary of recommendations of the Workshop participants is presented.

Changes in electrophysiological properties of cat hypoglossal motoneurons during carbachol-induced motor inhibition.

The control of hypoglossal motoneurons during sleep is important from a basic science perspective as well as to understand the bases for pharyngeal occlusion which results in the obstructive sleep apnea syndrome. In the present work, we used intracellular recording techniques to determine changes in membrane properties in adult cats in which atonia was produced by the injection of carbachol into the pontine tegmentum (AS-carbachol). During AS-carbachol, 86% of the recorded hypoglossal motoneurons were found to be postsynaptically inhibited on the basis of analyses of their electrical properties; the electrical properties of the remaining 14% were similar to motoneurons recorded during control conditions. Those cells that exhibited changes in their electrical properties during AS-carbachol also displayed large-amplitude inhibitory synaptic potentials. Following sciatic nerve stimulation, hypoglossal motoneurons which responded with a depolarizing potential during control conditions exhibited a hyperpolarizing potential during AS-carbachol. Both spontaneous and evoked inhibitory potentials recorded during AS-carbachol were comparable to those that have been previously observed in trigeminal and spinal cord motoneurons under similar experimental conditions as well as during naturally occurring active sleep. Calculations based on modeling the changes that we found in input resistance and membrane time constant with a three-compartment neuron model suggest that shunts are present in all three compartments of the hypoglossal motoneuron model. Taken together, these data indicate that postsynaptic inhibitory drives are widely distributed on the soma-dendritic tree of hypoglossal motoneurons during AS-carbachol. These postsynaptic inhibitory actions are likely to be involved in the pathophysiology of obstructive sleep apnea.

GABAergic neurons of the cat dorsal raphe nucleus express c-fos during carbachol-induced active sleep.

Serotonergic neurons of the dorsal raphe nucleus (DRN) cease firing during active sleep (AS, also called rapid-eye-movement sleep). This cessation of electrical activity is believed to play a 'permissive' role in the generation of AS. In the present study we explored the possibility that GABAergic cells in the DRN are involved in the suppression of serotonergic activity during AS. Accordingly, we examined whether immunocytochemically identified GABAergic neurons in the DRN were activated, as indicated by their expression of c-fos, during carbachol-induced AS (AS-carbachol). Three chronically-prepared cats were euthanized after prolonged episodes of AS that was induced by microinjections of carbachol into the nucleus pontis oralis. Another four cats (controls) were maintained 2 h in quiet wakefulness before being euthanized. Thereafter, immunocytochemical studies were performed on brainstem sections utilizing antibodies against Fos, GABA and serotonin. When compared with identically prepared tissue from awake cats, the number of Fos+ neurons was larger in the DRN during AS-carbachol (35.9+/-5.6 vs. 13.9+/-4.4, P<0.05). Furthermore, a larger number of GABA+ Fos+ neurons were observed during AS-carbachol than during wakefulness (24.8+/-3.3 vs. 4.0+/-1.0, P<0.001). These GABA+ Fos+ neurons were distributed asymmetrically with a larger number located ipsilaterally to the site of injection. There was no significant difference between control and experimental animals in the number of non-GABAergic neurons that expressed c-fos in the DRN. We therefore suggest that activated GABAergic neurons of the DRN are responsible for the inhibition of serotonergic neurons that occurs during natural AS.