Electrocorticogram in rats loaded with SART stress (repeated cold stress)

The electrocorticogram (ECoG) in a SART (specific alternation of rhythm in temperature)-stressed (repeatedly cold-stressed) rat, which is regarded as an experimental model for clinical vagotonic-type dysautonomia, was investigated in the present study by the power spectral technique. 1) Analysis of ECoG in SART-stressed rats during the resting-arousal state indicated a decrease in total power and a decrease in relative power in the delta band, and also an increase in relative power in the theta, alpha and beta bands. 2) In the slow-wave sleeping state, the ECoG of SART-stressed rats indicated a marked increase in total power, an increase in the delta band and decreases in theta, alpha and beta bands. 3) Electric stimulation of the posterior-hypothalamic area evoked alterations of ECoG similar to those caused by SART stress. ECoG response to electric stimulation in SART-stressed rats was less than that in unstressed rats. 4) Lesioning of the posterior-hypothalamic areas prevented SART stress-induced ECoG alterations. SART-stressed rats thus appear to be at a higher consciousness level on awakening but to sleep more soundly. They seem to exhibit greater fluctuation in brain activity than normal rats. There is also the possibility that the posterior-hypothalamic area is responsible to some degree for ECoG alterations in SART-stressed rats.     

Effect of 10 days reserpine or apomorphine administration on sleep cycles in rats.

The duration of sleep-wakefulness periods and sleep cycles was investigated by means of EEG and EMG in rats with implanted brain electrodes and myographic electrodes in neck muscles during 5 hr trials after repeated administration of 2 depressant drugs, reserpine and apomorphine. Ten days administration of reserpine at a dose of 0.05 mg/day, SC, decreased the duration of both slow-wave and paradoxical sleep, whereas apomorphine had no effect on sleep duration or occurance of sleep phases in either the low dose (0.2 mg/kg/day, SC) or in the high dose of 0.6 mg/kg/day, SC. This finding might be correlated with previously found differences in catecholamine levels in the brain in the same model experiments.     

Comparison between ethanol-induced and slow-wave sleep synchronous EEG activity utilizing spectral analysis

This study was designed to compare the cortical EEG and EEG power spectra associated with slow-wave sleep (SWS) with those after acute ethanol administration in the rat. Adult female Sprague-Dawley rats were implanted with chronic cortical and temporalis muscle electrodes and intravenous (i.v.) cannulas for drug administration. Ethanol was administered at doses of 1, 2 and 4 g/kg (i.v.) to 3 groups of 4 rats each, respectively. The EEG effects of ethanol were compared to those of slow-wave sleep. Ethanol produced dose-related increases in the mean duration of synchronous EEG episodes. Differences in EEG waveforms between the EEG in slow-wave sleep and ethanol-induced EEG synchrony were detected. Dose-related increases by ethanol in 0–4 Hz spectral power and decreases in 8–13 Hz spectral power were found. These results demonstrated discrete differences between the normal EEG in slow-wave sleep and ethanol-induced EEG synchrony in the rat.     

Eeg and behavioral effects of ethylketocyclazocine,morphine and cyclazocine in rats: Differential sensitivities towards naloxone

The effects of ethylketocyclazocine were compared with those of morphine and cyclazocine on the cortical EEG and spontaneous behavior of unrestrained, male Sprague-Dawley rats. Intraperitoneal administration of ethylketocyclazocine (2.5 mg/kg), morphine (10 mg/kg), and cyclazocine (2.5 mg/kg) each induced alterations in the EEG and behavior of rats characterized by high-voltage synchrony associated with behavioral stupor during wakefulness. Both ethylketocyclazocine and morphine produced biphasic EEG and behavioral profiles consisting of an initial phase of EEG synchrony and stupor followed by a period of EEG activation and behavioral arousal. In contrast, the profile of cyclazocine was not biphasic but was distinguished by intermittent periods of behavioral excitation and EEG desynchronization. When the rats were pretreated with naloxone (0.01–10 mg/kg, s.c. at ?10 min), there was a dose-related antagonism of the duration of EEG synchrony and stupor produced by the three analgesics. Ethylketocyclazocine was the most sensitive to naloxone (AD 50 dose of naloxone = 0.035 mg/kg) and cyclazocine was the least sensitive (AD 50 = 2.1 mg/kg). Antagonism of the drug-induced increase in latency to slow-wave sleep was dose-related for ethylketocyclazocine and morphine, but not for cyclazocine. The ability to identify differences between these three analgesics, based on their respective EEG and behavioral profiles and their sensitivities to naloxone, is consistent with current theories of multiple opiate receptors mediating certain pharmacological effects of opioids.     

Effects of the beta-carboline abecarnil on epileptic activity, EEG, sleep and behavior of rats.

The profile of the anxiolytic beta-carboline isopropyl 6-benzyloxy-4-methoxymethyl beta-carboline-3-carboxylate (abecarnil; ZK 112 119), a partial agonist at benzodiazepine receptors, was determined in two experiments. In the first, abecarnil was given to WAG/Rij rats; these rats generate spontaneously occurring spike-wave discharges and are regarded as a model for absence epilepsy. Effects were measured on epileptic activity, together with those on the spectral content of the background electroencephalograph (EEG), as well as on ongoing behavior. In a second experiment, effects on sleep and behavior were investigated in Wistar rats. It was found that, similarly to classical benzodiazepines, abecarnil possessed a strong antiepileptic character and also changed the background EEG to more high-frequency waves and less spindle activity. It also produced more immobile behavior. Abecarnil induced only small, marginally significant increases in slow-wave sleep while reducing REM sleep as a proportion of total sleep. It also reduced the number of REM periods. These observations are consistent with the proposed partial agonist activity of abecarnil, a drug with interesting therapeutic implications.     

Modulation of morphine-induced EEG and behavioral effects by dynorphin A-(1-13) in non-tolerant and morphine-tolerant rats

The purpose of the present study was to assess effects of dynorphin A-(1-13) on morphine-induced changes in electroencephalographic (EEG) spectral power and morphine-induced suppression of slow-wave sleep in non-tolerant and morphine-tolerant rats. Adult female Sprague-Dawley rats were implanted with chronic cortical EEG electrodes, electromyographic electrodes in the temporalis muscle and with intracerebroventricular (i.c.v.) cannulae and, in some cases, additional intravenous (i.v.) cannulae. Injections of morphine (i.c.v., 20 micrograms/rat) produced a biphasic EEG and behavioral response, composed of 2-3 hr of slow-wave bursts and increased spectral power (0-4 Hz) in the EEG, associated with behavioral stupor, followed by 2-3 hr of EEG and behavioral arousal. Dynorphin (i.c.v., 20 micrograms/rat), administered 10 min before injections of morphine in non-tolerant rats, antagonized morphine-induced increases in spectral power of the EEG and morphine-induced suppression of slow-wave sleep. In addition, EEG power spectra obtained after intraventricular administration of morphine from rats, treated with dynorphin and morphine intraventricularly 24 hr earlier, were qualitatively similar to those previously found after acute administration of kappa opioid agonists. In morphine-tolerant rats, pretreatment with dynorphin given intraventricularly, 10 min prior to intraventricular administration of morphine, restored morphine-induced increases in EEG spectral power and suppression of slow-wave sleep. The results suggest that dynorphin may modulate the characteristics of opioid receptors.     

Ultradian dynamics of sleep after a single dose of benzodiazepine hypnotics

A single bedtime dose of the benzodiazepine hypnotics, flunitrazepam (2 mg), triazolam (0.5 mg) or flurazepam (30 mg), was administered to young, healthy subjects. Abortive first rapid eye movement sleep (REMS) episodes, characterized by a low level of EEG slow-wave activity (spectral power density in the 0.75-4.5 Hz band) without rapid eye movements and/or muscle atonia, were more frequent in the drug night than in the placebo night or in the drug-free night following upon the drug night. The benzodiazepine hypnotics depressed slow-wave activity in non-REM sleep (NREMS) in the drug night and the subsequent drug-free night. However, the typical declining trend of slow-wave activity over the first three NREMS-REMS cycles, and the cyclic ultradian pattern of slow-wave activity were little affected by the hypnotics. The results indicate that benzodiazepine hypnotics depress the generation of slow EEG waves without disrupting the homeostatic and ultradian processes of sleep regulation.     

Effects of acute delta 9-THC administration on EEG and EEG power spectra in the rat

This study was designed to determine the acute effects of delta 9-THC on the cortical EEG with the spectral analysis technique. Adult female Sprague-Dawley rats were implanted with chronic cortical and temporalis muscle electrodes. Intraperitoneally administered delta 9-THC (5 and 10 mg/kg) produced a reduction in peak-to-peak voltage of the desynchronized cortical EEG during wakefulness. Associated spectral power was reduced to about 50% of control during the first hour after injection of delta 9-THC and gradually returned toward the control value over an 8-hr period. Occurrences of delta 9-THC-induced high-voltage EEG bursts, overriding the reduced EEG tracing, were associated with an EEG spectral peak at 6 Hz. The first few slow-wave sleep (SWS) episodes appearing after delta 9-THC administration were associated with more slow-frequency waveforms and more slow-frequency spectral power than with control slow-wave sleep episodes. During control rapid eye movement (REM) sleep episodes, an EEG theta wave pattern, with an associated spectral peak at about 8 Hz, was characteristic. Conversely, the first few REM sleep episodes emerging after delta 9-THC administration contained overriding high-voltage bursts, the related power spectra of which had two peaks at about 7 and 11 Hz.     

Effect of tryptophan on sleep in the rat

Rats implanted with EEG and EMG electrodes were injected intraperitoneally with 30 or 120 mg/kg of l-tryptophan and a recording was made for 6 hr. The 30 mg/kg dose decreased the latency to the first slow-wave sleep (SWS) episode by 15 min (43%) and to the first rapid eye movement (REM) sleep episode by 18 min (27%). There were no changes in wakefulness. SWS or REM sleep during the 6hr of EEG recording. In contrast, the 120 mg/kg dose did not affect sleep latencies but increased wakefulness and decreased REM sleep during the second hour of EEG recording.The reduction of SWS latency produced by the 30 mg/kg tryptophan occurred at a time when the level of 5-hydroxyindoleacctic acid (5-HIAA) was elevated in the cortex and pons-medulla (i.e. 15 min post-injection). At the same time, hippocampal and cortical dopamine levels decreased while that of homovanillic acid did not change. The level of norepinephrine, also, decreased in the hippocampus. Whereas these changes in brain calecholamines were not observed 45 min after tryptophan administration, there was an elevation of 5-hydroxytryptamine in the pons-medulla and cortex as well as of 5-HIAA in the pons-medulla and hippocampus.Present data suggest that the hypnotic effect of tryptophan may involve the normal sleep mechanism, since similar neurochemical findings were reported during natural SWS. However, the effect may be limited to a certain dose range because a high dose produced waking rather than sleep.     

Effect of chronic treatment with milnacipran on sleep architecture in rats compared with paroxetine and imipramine

A number of studies in humans and various other species have shown that chronic treatment with antidepressants, such as tricyclics or selective serotonin reuptake inhibitors (SSRIs), induces a decrease or suppression of rapid eye movement (REM) sleep. The effect of a new selective serotonin and noradrenaline reuptake inhibiting (SNRI) antidepressant, milnacipran, on REM sleep has been investigated and compared with that of the SSRI, paroxetine, and the tricyclic, imipramine. Rats injected with vehicle or milnacipran twice a day showed, over 24 h, a similar amount of REM sleep, number and duration of REM sleep episodes to control rats. In contrast, rats treated acutely with imipramine or paroxetine showed a statistically significant decrease in the total quantity of REM sleep. The number of REM sleep episodes was decreased while their duration was increased. A more detailed analysis showed further that the quantity of REM sleep was decreased for the first 4 h following the 9 a.m. injection but not the 7 p.m. injection for milnacipran, during the first 6 h for paroxetine and for the entire light-dark period for imipramine. For all drugs, the quantities of slow-wave sleep and waking over 24 h were not significantly different from control conditions and no rebound of REM sleep occurred during the day following withdrawal. Power spectrum analysis revealed no global changes in the different electroencephalogram (EEG) waves (delta, theta, gamma) between the control condition and the different treatments during waking, slow-wave sleep or REM sleep. Taken together our results indicate that the SNRI, milnacipran, at therapeutic doses, induces only minor disturbances of REM sleep compared with a SSRI and tricyclic antidepressant used. Possible mechanisms responsible for the difference of action on REM sleep of milnacipran are discussed.     

Chlorpromazine and human sleep

The aim of this study was to examine human physiological sleep profiles, including the amount and distribution of electroenoephalographic (EEG) stages of sleep, variations in specific frequency bands in the EEG spectrum and certain phasic phenomena such as movement arousals, sigma spindles and rapid eye movements, following oral administration of a moderate dose (150 mg) of chlorpromazine (CPZ) to 12 young male volunteers. At this dose level the drug had few systematic effects on sleep, although it did reduce the latency of onset of stage REM and the number of movement arousals, while increasing the amount of slow-wave (SW) sleep. These effects persisted during the post-medication recovery night, but at no time was there any systematic change in the total amount or percent of REM sleep, the duration of the REM-to-REM cycle, the average length of REM episodes or the density of rapid eye movements during stage REM. Frequency analysis of EEG revealed that CPZ produced a trend toward increased fast (beta) activity recorded from pre-central placements during stage REM, and reduced density of sigma spindles in stage 2 sleep. Thus, for the most part, a single moderate dose of CPZ left the tonic, phasic and sequential properties of the sleep cycle unaltered. These results confirm previous investigations showing that for small to moderate clinical doses, CPZ invariably enhances SW sleep and reduces the frequency of movement arousals. On the other hand, the effect of the drug on stage REM apparently depends on dose. Small doses potentiate REM sleep or accelerate its onset, whereas larger doses either reduce stage REM or leave it unaffected. Several authors have pointed out that most hypnotic agents cause substantial alterations of the sleep profile, and that their withdrawal can cause profound disruption of sleep and marked clinical disturbance. It also has been suggested that there exists a relation between drug dependency and the degree of initial REM suppression caused by a drug. The finding confirmed by the present study that clinical doses of CPZ cause mild sedation, and enhanced SW sleep without any significant modification of REM, sleep, indicates that CPZ has features which may recommend it as a standard hypnotic.This study was supported in part by USPHS grant No. MH 10844-05 of the National Institute of Mental Health.We thank David Paskewitz and O. H. Rundell, Jr., for their assistance in design and management of the study and Rosa Coulter for her contributions to data analysis.     

Comparative studies on antidepressant action of alprazolam in different animal models

Alprazolam, a new benzodiazepine from triazolobenzodiazepine group, produced anxiolytic action in the conflict test with potency similar to that of diazepam. The myorelaxant activity of the drug was relatively weak. Unlike desipramine, alprazolam failed to reduce the immobility of rats in the forced swim test and was unable to prevent clonidine-induced hypothermia. Alprazolam, unlike desipramine, failed also to potentiate behavioral effect of noradrenaline injected into the hippocampus. Alprazolam after acute but not chronic administration antagonized the synchronizing effect of clonidine on EEG pattern. On the other hand, alprazolam similarly to tricyclic antidepressants, prevented the suppression of dominance behavior by clonidine in rats competing for food. The results indicate that alprazolam acts only weakly upon noradrenergic mechanisms related to depression and to antidepressant action of drugs.     

Sleep induced by drugs injected into the inferior horn of the lateral cerebral ventricle in dogs.

1 In unanaesthetized dogs, cholinomimetic drugs and their antagonists were injected into the inferior horn of the left lateral cerebral ventricle. Injection volumes of 5 mul were used to limit spread of the drugs beyond the inferior horn. The effects on EEG and behaviour were recorded and compared with the effects of the same doses given into the body of the right lateral ventricle a little behind the foramen on Monro. 2 injections of cholinomimetic drugs into the inferior horn (acetylcholine 1-2 mug, physostigmine 1.0 mug, pilocarpine 100 mug and nicotine 10 mug) induced sleep during the following hour. The same doses injected into the body of the lateral ventricle did not produce sleep. 3 Cholinolytic drugs (atropine 10-20 mug, hyoscine 0.4-1.6 mug (+/-)-tubocuraine 10-20 ng and hexamethonium 40 mug) injected into the inferior horn also produced sleep, but the same doses injected into the body of the lateral ventricle were without effect. The EEG recorded after tubocurarine showed high voltage slow waves during sleep and desynchronized activation during rapid eye movement sleep. 4 Noradrenaline (10 mug) injected into the inferior horn produced sleep whereas the same dose given into the body of lateral ventricle did not produce sleep. The results with 5-hydroxytryptamine were equivocal. 5 It is suggested that the site for induction of sleep lies in structures lining the inferior horn of the lateral cerebral ventricle and that the cholinomimetic drugs probably act by a depolarizing block and the acetlycholine antagonists by a competitive block.     

Secobarbital and nocturnal physiological patterns

Summary Sleep patterns of 14 male Ss were examined following a single oral dose (200 mg) of the barbiturate secobarbital. Compared to baseline, medication caused definite changes in the amount and distribution of the EEG stages of sleep. With the drug, a decrease in percent stage REM and an increase in percent stage 2 were accompanied by fewer body movements and a trend toward less waking. A more striking effect was the drug-induced redistribution of EEG stages with slow-wave sleep potentiated during the first half of the night but virtually eliminated during the last half. Stage REM, on the other hand, was inhibited in the first half but returned to baseline levels in the last half of the night. Recent evidence suggests that such effects could result from modulation of brain levels of the monoamines.Within the stages of sleep the amount of fast EEG activity was increased by the drug, with a tendency toward desynchrony. Pre-central beta activity became especially prominent in stages REM and 2 (low-voltage phases), and this change was associated with inhibition of such phasic events as eye movements during REM sleep, sigma spindles during stage 2 and spontaneous electrodermal responses in slow-wave sleep. Thus, it appears that secobarbital potentiates tonic and suppresses phasic phenomena during sleep. A possible interpretation of these results is that secobarbital enhances electrical activity in forebrain structures (Routtenberg's arousal system II) while inhibiting the reticular activating system (arousal system I) causing a reduction of phasic variability in all stages of sleep.We wish to thank O. H. Rundell and Joe Gold for their assistance in data acquisition and management of the experiment, and Rosa Coulter, Cindy Williams and Maria Chan for their invaluable contributions to data reduction and statistical analysis.     

Naltrindole retards tolerance development to morphine-induced effects on EEG and EEG power spectra.

In the present study, EEG and EEG power spectra were used to assess the effects of naltrindole, a selective delta opioid antagonist, on the development of tolerance to morphine. Adult female Sprague-Dawley rats were implanted with cortical EEG electrodes and permanent indwelling i.c.v. and i.v. cannulas. Twice daily for 7 days, rats were pretreated with either i.c.v. naltrindole (20 nmol) or i.c.v. water, 20 min before i.v. morphine (10 mg/kg) injections. The treatments produced EEG slow-wave bursts and associated behavioral stupor. The amount and duration of these effects decreased less rapidly over the 7 days in the naltrindole-pretreated rats than in the water-pretreated rats. I.c.v. naltrindole pretreatment also prevented significant decreases in latency to onset of slow-wave sleep that were seen in the i.c.v. water-pretreated group. EEG data were further analyzed on a Pathfinder II computer. The development of tolerance was reflected by decreases in the total absolute EEG spectral power (1-50 Hz) over the 7-day period. Rats that were pretreated with i.c.v. naltrindole (20 nmol) did not display a significant decrease in total absolute EEG spectral power by the 7th day, as did the i.c.v. water-pretreated group. Furthermore, significant differences were seen for complexity, mobility, and edge frequency between the two pretreatment groups. A delayed qualitative change in the EEG power spectra was also observed in rats pretreated with i.c.v. naltrindole. On day 1, EEG slow-wave bursts were associated with increases in EEG spectral power over the 1-10 Hz range.(ABSTRACT TRUNCATED AT 250 WORDS)     

Changes in the rat sleep-wake cycle produced by DL-6-fluorotryptophan,a competitive inhibitor of tryptophan hydroxylase

DL-6-Fluorotryptophan (6-FT), a competitive inhibitor of tryptophan hydroxylase, produced a transient disruption of sleep in rats chronically implanted with EEG recording electrodes. In the 4 h period following the administration of 6-FT (120 mg/kg) awake time was increased, paradoxical sleep time was decreased and slow-wave sleep remained unchanged. These sleep changes were accompanied by significant reductions in brain 5-HT levels. L-Tryptophan (100 mg/kg) co-administered with 6-FT prevented the major sleep changes whereas L-leucine (100 mg/kg) was without effect. The major sleep changes produced by 6-FT were prevented by the pineal indole melatonin (20 mg/kg) but not by L-5-hydroxytryptophan (5 mg/kg). These neurochemical and drug interaction data raise the possiblity that 5-hydroxytryptamine is involved in the control of paradoxical rather than slow-wave sleep in the rat.     

药源性睡行症的常见致病药物及防治

睡行症是一种在睡眠中出现的以行走或其他异常行为或活动为特征的睡眠障碍（sleep disorder），通常发生在非快速眼动睡眠的慢波期。由药物引起的睡行症称之为药源性睡行症。引起药源性睡行症的常见药物有镇静催眠药、抗精神病药物以及抗抑郁症药等。据称，药源性睡行症的发生由多因素所致，包括既往睡行症发作史，应用增加慢波睡眠的药物以及体内外刺激。药源性睡行症的发生机制尚不清楚，有人认为是某些神经递质如5-羟色胺、1-氨基丁酸增加慢波睡眠所致。药源性睡行症防治原则包括：减少致病药物的剂量或停药，排除危险因素，加强环境安全及给予苯二氮[艹卓]革类药物。     

Acute cortisol administration promotes sleep intensity in man.

The neuronal mechanisms of sleep generation, in particular synchronization of brain activity in the process of non-rapid-eye movement (non-REM) sleep, has been elucidated in the past decade. A previous study of our group showed that acute administration of cortisol is known to increase slow-wave sleep and suppress rapid-eye movement (REM) sleep in man. To further elucidate the non-REM sleep-promoting effects of cortisol with respect to the synchronization of cortical activity, it is important to establish a sleep-state-specific quantitative EEG analysis. We therefore investigated the effects of repetitive injections of hydrocortisone on spectral composition of sleep EEG in 10 healthy male young volunteers. In addition, we performed high-frequency blood samplings to assess the relation between changes in the sleep EEG and sleep-associated secretion of growth hormone (GH). Cortisol administration resulted in a significant increase in highly synchronized EEG activity including delta and theta frequencies, according to a higher amount of slow-wave sleep. This effect predominated in the first few hours of night sleep. REM sleep was decreased, which appeared to be secondary to the lengthened first sleep cycle. The cortisol-induced stimulation of GH release did not occur in correspondence with the increased slow-wave activity. In view of the sleep impairing properties of corticotropin-releasing hormone (CRH) and the sleep-promoting function of GH-releasing hormone, it appears likely that a negative feedback inhibition of endogenous CRH was the key mechanism mediating the observed results. The cortisol-induced effects on sleep intensity and sleep-associated GH secretion appeared to be driven by different mechanisms.     

Electroencephalographic study on the central action of a new anxiolytic: 3a alpha, 4 beta, 7 beta, 7a alpha-hexahydro-2-(4-(4-(2-pyrimidinyl)-1- piperazinyl)-butyl)-4, 7-methano-1H-isoindole-1, 3(2H)-dione dihydrogen citrate (SM-3997)

Electroencephalographic (EEG) studies were performed to examine the effects of SM-3997 on the spontaneous EEG, EEG arousal responses, recruiting responses and hippocampal afterdischarges in rabbits and the spontaneous EEG in chronically electrode-implanted rats. In acute experiments using rabbits, SM-3997 at doses of 1-3 mg/kg, i.v., produced low-voltage fast waves in cortical EEG and slow waves with reduction of the amplitude in hippocampal EEG. The drug at doses of 1-3 mg/kg, i.v., dose-dependently inhibited the threshold stimulus voltages in EEG arousal responses induced by stimulation of the midbrain reticular formation and slightly inhibited the threshold in recruiting responses by stimulation of the centromedian nucleus of the thalamus. However, the cortical and hippocampal afterdischarges induced by hippocampal stimulation remained unaffected by SM-3997 at doses up to 3 mg/kg, i.v., while they were inhibited by diazepam of 1 mg/kg, i.v. In the study using rats in which electrodes were chronically implanted, SM-3997 at doses of 10-30 mg/kg, i.p., also produced low voltage fast waves in cortical EEG and slow waves of reduced amplitude in hippocampal EEG; and it simultaneously caused flat body posture. These results suggest that SM-3997 acts on both the cerebral cortex and hippocampus, inducing much more pronounced inhibition on the midbrain reticular formation-hippocampal system     

Effects of N-cyclopentyladenosine and caffeine on sleep regulation in the rat

To study the role of adenosine in sleep regulation, the adenosine A₁ receptor agonist N⁶-cyclopentyladenosine (CPA) and the antagonist caffeine were administered to rats. Intraperitoneal (i.p.) CPA 1 mg/kg but not 0.1 mg/kg, suppressed rapid-eye-movement (REM) sleep and enhanced electroencephalographic (EEG) slow-wave activity (power density 0.75–4.0 Hz) in non-REM sleep. The latter effect was remarkably similar to the response to 6-h sleep deprivation. The effects persisted when CPA-induced hypothermia was prevented. Caffeine (10 and 15 mg/kg i.p.) elicited a dose-dependent increase in waking followed by a prolonged increase of slow-wave activity in non-REM sleep. The combination of caffeine (15 mg/kg) and sleep deprivation caused less increase in slow-wave activity than sleep deprivation alone, indicating that caffeine may reduce the buildup of sleep pressure during waking. The results are consistent with the involvement of adenosine in the regulation of non-REM sleep.