Previous studies have shown that the region of the nucleus of the solitary tract (NST) is involved in the control of electrocortical activity and in sleep mechanisms. It also is well known that this region contains the highest concentration of opioid receptors within the medullary brainstem. The involvement of the NST opioid system in sleep-wakefulness states were evaluated.
Ten cats were implanted with electrodes for chronic polygraphic recordings of their sleep-wakefulness states and provided with an implanted guide cannula stereotaxically aimed at the NST region. Microinjections of saline, morphine sulfate, morphiceptin (specific mu agonist), D-pen-2-D-pen-5-enkephalin (delta agonist), and U-50488H (kappa agonist) were given to the freely moving animals (doses 0.8-2.4 x 10(-9) M, in a volume of 0.05 microliters of saline). After microinjections, sleep-wakefulness recordings were obtained for 8 h.
Morphine microinjections in NST provoked a dose-dependent enhancement of all the polygraphic and behavioral manifestations of slow wave sleep. This effect was blocked by the prior intraperitoneal administration of naloxone. The mu and delta agonists also produced a hypnotic effect by enhancing slow wave sleep. By contrast, the kappa agonist caused no changes in sleep-wakefulness states.
These results indicate that endogenous opioids could be involved in controlling electrocortical activity generated by NST and that activation of mu and delta NST opioid receptors enhanced the electroencephalographic synchronization associated with behavioral slow wave sleep in cats.
Key words: Brain, brainstem: nucleus of the solitary tract. Cat. Analgesics, opioid; delta-receptor agonist; kappa-receptor agonist; morphine; micro-receptor agonist. Sleep: electroencephalogram; slow wave sleep.
NATURALLY occurring opioids have been used for many thousands of years for their ability to depress the level of consciousness. Their strong analgesic effects were recognized much later. Although the central nervous system (CNS) structures and the opioid receptors involved in this analgesic effect have been identified, the sites of action and the opioid receptors that mediate other CNS effects of opioids, such as behavioral changes and variations in the electroencephalogram (EEG) and in the sleep-wakefulness cycle (SWC) remain unclear. The need to understand the mechanisms of opioid actions in the CNS in relation to the latter actions is important in anesthesiology given the EEG and sleep disturbances produced by opioids during intraoperative and postoperative periods. [1,2] Additionally, the problem of intraoperative recall and awareness under general anesthesia often has been reported,  and the drowsiness and decreased awareness commonly observed after opioid administration are considered adverse effects in pain treatment.  Furthermore, the incidence of perioperative cardiopulmonary complications associated with the sleep state in patients who have received opioids is also significant. .
The behavioral effects provoked by opioids are typically complex because excitatory and inhibitory behavioral patterns coexist in humans and in different animal species. Systemic administration of morphine suppresses both slow wave sleep (SWS) and paradoxical sleep (PS)[6–8]; however, animal studies during the insomnia period have documented that opioids cause a typical EEG/behavioral dissociation, i.e., high-voltage slow-frequency waves are present in the EEG while the animals remain awake exhibiting stupor [7,9,10] or other more active behavior. [8,11].
A number of experimental findings have indicated that the interaction of morphine with the forebrain has an excitatory effect on behavior  that is associated with EEG desynchronization, [12,13] this in contrast to morphine's effects in the whole lower brainstem, where, with a dose-dependent decrease of PS,  it produces behavioral signs of SWS  and EEG synchronization.  Recent experiments using microinjection techniques  have indicated that local administration of morphine in the medial pontine reticular formation inhibits both SWS and PS. However, little is known about other specific brainstem structures and receptors that could be implicated in the excitatory and/or inhibitory behavioral and EEG effects produced by opioids.
The density of opioid receptors [16–18] as well as neuronal enkephalinergic endings and somata  in the region that Ramon y Cajal  called complex of the solitary tract, located in the floor of the fourth ventricle, is very high. The visceral sensitive information from the 7th, 9th, and 10th cranial nerves is organized in the region of the nucleus of the solitary tract (NST), which mediates the most important respiratory  and cardiovascular  reflexes. In addition, it has long been recognized that electrical stimulation of the NST produces EEG synchronization,  and more recent experimental data confirmed the important hypnogenic and EEG synchronizing role of this region. [24–27].
These facts led us to evaluate the possible involvement of NST opioid receptors in both EEG synchronization and behavioral depression. To test this hypothesis, morphine sulfate and highly specific agonists of micro, delta, and kappa opioid receptors were unilaterally microinjected into the NST of intact, unanesthetized cats and, after the microinjections, their SWC states were monitored. The results revealed that administration of morphine, or of either of the specific micro and delta agonists in the NST, enhanced EEG synchronization associated with SWS behavior.
Material and Methods
After institutional approval was obtained, ten adult cats were anesthetized with sodium pentobarbital (35 mg *symbol* kg sup -1 intraperitoneally) and, under sterile conditions, implanted with electrodes for chronic polygraphic recordings of the SWC states. These electrodes recorded EEG, electroculogram, electromyogram, and ponto-geniculo-occipital waves from the lateral geniculate bodies of the thalamus. As part of the same surgery, a 20-G cannula provided with a stylet was implanted, stereotaxically aimed at the NST according to the coordinates of the Reinoso-Suarez' atlas (AP - 10.0, V 3.5, and L 3.0, theta = 18 degrees from vertical). The tip of the cannula was left 4.0 mm above the target site. All the electrodes were terminated in an Amphenol connector. The electrodes and the cannula were secured permanently with acrylic cement to the skull of the animal.
Eight to 10 days after surgery, each cat was acclimatized for 2–3 days to the recording conditions. The animal was placed in a sound-attenuated ventilated recording chamber, with food and water ad libitum, where it could be observed through a one-way mirror. After becoming familiar with the chamber, all cats had a baseline polygraphic recording before starting the treatments with opioid microinjections in the NTS area.
When the cat was awake, but gently restrained in a loosely fitting bag, morphine sulfate, morphiceptin (specific micro agonist ). D-pen-2-D-pen-5-enkephalin (DPDPE, a highly selective delta agonist ) or U-50488-H (kappa agonist ) in a volume of 0.05 micro liter of saline was injected into the NST using a 0.5 micro liter Hamilton syringe inserted through the stereotaxically implanted cannula. The tip of the needle protruded 4 mm from the cannula. All cats received single equimolar microinjections of each opioid agonist separated by an interval of at least a week from the next microinjection. Six cats received a dose of 1.7 x 109M of each opioid agonist (three of them received additional treatments with the lowest dose 0.8 x 109M of each agonist); the remaining four cats received the high dose (2.4 x 109M) of each opioid agonist. In all cats, control data were collected using identical procedures after microinjecting 0.05 micro liter of saline in the NST area. Without strictly following a Latin square design, opioid and saline microinjections were given in a quasirandomized order. Finally, in six cats in which opioids had been administered previously, microinjections of morphine sulfate (2.4 x 109M in three cats and 1.7 x 109M in the remaining three) were repeated 10 min after administering intraperitoneal naloxone 2 mg/kg. These combined morphine-naloxone experiments were done at the end of the series of opioid treatments in four animals and in middle of the series in the remaining two animals. After receiving the microinjections, the cats were released in the sound-attenuated chamber, and the polygraphic recordings were immediately begun. Baseline and experimental polygraphic recordings lasted at least 8 h. Behavioral observations and videotape recordings were made during the first 2 h of the recordings.
Since behavioral observations indicated that none of the cats presented any EEG/behavioral dissociation after opioid delivery in the NST region, polygraphic recordings were scored visually in 1-min epochs according to the polygraphic criteria of wakefulness, drowsiness, SWS, and PS of the cat. [11,32] Wakefulness is characterized by desynchronized EEG with low voltage/fast activity, well sustained electromyogram, and frequent eye movements; drowsiness by synchronized EEG with bursts of high voltage waves superimposed over low voltage/fast activity, well sustained electromyogram and occasional eye movements; SWS by synchronized EEG with continuous high voltage/slow activity, diminished electromyogram activity and no eye movements; and PS by desynchronized EEG with low voltage/fast activity, ponto-geniculo-occipital waves, muscle atonia and rapid eye movements. The sleep recordings were blocked in 1-h intervals. The percentage of time spent in each SWC state and the mean duration and the number of its episodes were calculated for each 1-h interval. All statistical analyses were based on the values of the l-h blocks after baseline and saline controls and the different opioid treatments.
According to the dose and the location of the microinjections, the cats were classified into three groups (Results). In each group of cats, changes in wakefulness, drowsiness, SWS, and PS among baseline, saline, and the different opioid treatments during either the entire 8 h or the first 4 h of recordings were assessed by one-way analysis of variance for repeated measures. Further multiple pair contrasts were made using Fisher's pairwise comparison test. Statistical significance was set at P less or equal to 0.05.
After the experiments were finished, the cats were deeply anesthetized with sodium pentobarbital. Brains were perfused by cannulating the aorta and infusing saline and a 10% formalin solution. Brains were frozen, sectioned at 50 micro meter thickness, and stained with Nissl stain. Each frontal section containing the needle tract was projected on sagittal planes of the cat's brain.
Behavioral observations during the polygraphic recordings indicated that there was a strict correlation between behavior and the polygraphic events that characterize the different SWC states of the normal intact cat in all our experiments. Local microinjections of morphine and the other opioids delivered in the lower brainstem did not produce the typical EEG/behavioral dissociation observed after systemic administration of low doses of morphine in intact cats [8,11] and in other species. [7,9].
The photomicrograph of Figure 1shows the injection site in one animal whose microinjections were located in the NST region, and Figure 2shows the sequential organization of the SWC states in representative experiments in the same cat during the first 2 h of recordings. After a saline microinjection in the NST (saline-NST, Figure 2) the different states of the SWC alternated one after another. However, when morphine was microinjected in the NST (morphine-NST, Figure 2), the cat exhibited a long episode of SWS with all its behavioral and polygraphic characteristics. Further facilitation of SWS episodes also took place in the second hour of the recording. However, this SWS enhancement did not occur when intraperitoneal naloxone was administered before the microinjection of morphine into the NST (intraperitoneal naloxone plus NST-morphine, Figure 2).
Histologic analyses revealed that the microinjections were located within the NST region in seven cats, whereas in the three remaining cats, the microinjections were situated in neighboring lower brainstem structures. The location of all the microinjection sites is summarized schematically in Figure 3. The cats were divided into three groups depending on the location of the microinjection and the doses of the opioids: group A, four cats with NST microinjections of which received the highest (2.4 x 10 sup - 9 M) opioid dose; group B, three cats with NST microinjections of 1.7 x 10 sup -9 M of opioids (one animal on this group also had the treatments with the lowest opioid dose (0.8 x 10 sup -9 M)); group C: the remaining three cats with 1.7 x 10 sup -9 M of opioid microinjections located outside the NST region (two animals of this group also had microinjections at the 0.8 x 10 sup -9 dose). Because analyses of variance for repeated measures were used in three groups to determine statistically significant changes in each state of the SWC, data from the 0.8 x 10 sup -9 M experiments were excluded from groups B and C.
In group A, analyses of variance showed that there were significant variations among treatments in the amount of time spent in both SWS (F5,155 = 9.127, P < 0.001) and wakefulness (F5,155 = 4.158, P < 0.001) during the entire 8 h of recordings; changes in drowsiness and PS did not reach statistically significant values (time for drowsiness, F5,155 = 2.271, P < 0.10; for PS, F5,155 = 1.839, P < 0.20). Pairwise comparisons (Figure 4) indicated that none of the SWC states presented statistically significant variations in baseline and saline control experiments, however, the effects of NST administration of 2.4 X 109M morphine, morphiceptin or D-pen-2-D-pen-5-enkephalin (DPDPE) significantly increased the amount of SWS as compared to baseline or saline controls, and the increase in SWS occurred mainly at the expense of wakefulness, which was statistically decreased with the same opioid treatments (Figure 4). NST microinjections of U-50488-H did not provoke any statistically significant change in the proportions of the SWC states as compared with the baseline and saline controls (Figure 4). As shown in Table 1, the enhancement of SWS produced by the morphine microinjections in the NST was produced by a significant increase in the mean duration of its episodes (F2,62 = 3.694, P = 0.03) and not in the number of the SWS episodes (F2,62 = 1.249, P = 0.294). The mean duration and the number PS episodes after administration of morphine in the NST did not significantly differ from the values of the baseline (F2,62 = 0.508, P = 0.604) or saline (F2,62 = 0.379, P = 0.686) control recordings.
In group B, changes in the proportions of wakefulness, drowsiness, SWS, and PS did not reach statistically significant levels for the entire 8 h of recordings, as shown by the corresponding analyses of variance (time for wakefulness, F5,115 = 1.315, P < 0.30; for drowsiness, F5,115 = 1.281, P < 0.30; for SWS. F5,115 = 2.184, P < 0.10; and for PS F5,115 = 1,219, P < 0.40). However, the different treatments produced significant variations in the amount of time spent in SWS and drowsiness during the first 4 h of recordings (Table 2). Changes in wakefulness and PS did not reach statistically significant differences (Table 2). Pairwise comparisons indicated that the effects of the administration of 1.7 X 10 sup -9 M of morphine, morphiceptin, or DPDPE in the NST significantly increased the time spent in SWS during the first 4 h as compared with baseline or saline-injected controls. These substances also decreased the time spent in drowsiness, but these changes only were significant in comparison with baseline controls. In this group of cats, NST microinjections of U-50488-H did not change proportions of the SWC states when compared with the proportions in baseline and saline controls.
In the three group C animals, in which the microinjections did not reach the NST, during the whole 8 h of recording or during the first 4 h of recording (Table 2), wakefulness, drowsiness, SWS, and PS were unaffected after the different treatments.
(Figure 5) shows the amount of time spent in SWS during the first hour of recordings in combined experiments of naloxone intraperitoneally administered 10 min before morphine delivery in the NST in three cats from group A. In these experiments, the amount of SWS did not significantly differ from that found in the saline controls; furthermore, in the morphine-naloxone experiments, these animals spent significantly less time in SWS than they did after NST administration of morphine alone. Thus, pretreatment with naloxone effectively blocked the increment of SWS produced by morphine-delivered NTS.
Parenteral administration of clinical doses of opioids provokes drowsiness and decreases the level of consciousness in humans. However, EEG and polygraphic studies after parenteral administration of morphine in humans contradictorily reveal both a decrease in SWS sleep (stages III and IV) and PS and an increase in wakefulness and stage I sleep.  Likewise, both excessive sedation or behavioral excitation, with different time courses, can be observed after intrathecal administration of morphine.  Studies in other species have confirmed the sleep-suppressing effects of systemic or intraventricular morphine [7,8,11] and other opioids. [34–36] In addition, these studies have indicated that, during this insomnia, there is an EEG/behavioral dissociation: the desynchronized EEG pattern typical of normal wakefulness is usually absent; instead, behavioral wakefulness is accompanied by a synchronized EEG with high-voltage slow waves resembling the pattern that appears during physiologic SWS.
Several factors might explain the complex excitatory/inhibitory actions evoked by opioids, namely: dose dependency, the stimulation of different types of opioid receptors, and different sites of interaction within the CNS. Although experimental studies in different animal species have demonstrated hyperactive or depressive behavioral states respectively associated with high and low doses of morphine, [37,38] this duality in the effects does not depend exclusively on the dose, because studies in rats  and cats [8,11] have shown that low doses of systemic morphine provoke the EEG/behavioral dissociation typical of opioids. It is also unlikely that the excitatory/inhibitory actions depend on different specific types of opioid receptors, because the administration of more selective agonists of opioid receptors, such as enkephalin analogs  or cyclazocine,  still provokes EEG/behavioral dissociation.
The current study demonstrates that microinjection of morphine and other agonists of micro- and delta-opioid receptors into the NST increases SWS. No EEG/behavioral dissociation was observed in our experiments when small microinjections of morphine and other opioids were locally delivered in the NST region or in neighboring structures, because we always observed that fully synchronized EEG was associated with behavioral sleep and with other polygraphic features that characterize SWS. The opioid-induced increase of SWS seen in the current study was site-specific. The hypnogenic and EEG synchronizing action occurred only when morphine and the micro and delta opioids were administered in the NST region. Estimates based on microinjections of dyes in the NST of a volume similar to the one used in our study (50 nl) have indicated that the initial area of diffusion has a long diameter, about 500 micro meter.  In our experiments, morphine and the micro and delta opioids were unable to increase SWS when microinjected into sites only 1 mm away from NST (Figure 3).
Previous studies dealing with the sites of action of morphine for generation of EEG and SWC effects have reported a variety of results depending on the region of the CNS that received the stimulation. EEG and behavioral activating effects mediated by the prosencephalon have been described in morphine-treated animals with occlusion of the aqueduct of Sylvio  or with a brainstem transection ; in contrast, synchronized EEG and sleep behavior seem to be the result of the interaction of the drug with the whole lower brainstem. [12,14] The role of specific brainstem structures other than the NST in mediating sleep opioid effects has been examined previously using the microinjection technique. Keifer et al.  reported that morphine locally administered in the medial pontine reticular formation promotes behavioral and electrographic wakefulness with inhibition of both SWS and PS; however, morphine microinjections in the dorsolateral pontine tegmentum at the level of the locus ceruleus increased SWS.  Therefore, it is evident that either activating or depressing actions can be obtained depending on the CNS loci at which the opioids act. All of these results emphasize the importance of determining the brain sites through which opioids alter EEG and behavioral arousal.
Interspecies differences between cats and humans partially limit the extrapolation of the results of the current study on behavioral effects and EEG changes associated with perioperative use of opioids to humans. However, the EEG and sleep responses of cats [8,11] to relatively low doses of morphine are similar to those found in humans and other species. These findings, along with the wide use of the cat in the research of sleep mechanisms, support the validity of this species as an experimental model for the study of the neural mechanisms involved in EEG and sleep effects of opioids. Thus, the complex and often contradictory behavior observed in humans [6,32] under systemic morphine, as well as the EEG/behavioral dissociation described in animals, [7,8] could be explained by the simultaneous activation of both EEG-desynchronizing and EEG-synchronizing structures. Also, the perioperative hypnosis and EEG synchronization induced by small doses of opioids, when administered with other anesthetic agents,  could be explained by the activation of synchronizing structures like the NST region and the inhibition of the CNS-activating mechanisms by the anesthetic drugs. What is more, the very high doses of micro-opioid agonist required for adequate intraoperative hypnosis [42,43] could be explained by the need for predominant stimulation of micro receptors located in synchronizing structures. On the other hand, because interaction of morphine with NST lengthens the SWS episodes, opioids also would strengthen the hypnotic effects of anesthetic drugs  by enhancing the duration of the SWS episodes once these were induced by the anesthetic agents. This would explain why the drowsiness after administration of parenteral opioids can be strongly exacerbated by the concomitant use of any hypnotic drug, even in small doses, or by the presence of a physiologic period of sleep, independently of adequate therapeutic plasmatic concentration of the opioid drug employed. .
In our experiments, intraperitoneal pretreatment with naloxone effectively blocked the SWS increase produced by the NST morphine microinjection. Pretreatment with the antagonist blocked the sleep-suppressing effects of morphine in the medial pontine reticular formation.  These findings are consistent with the hypothesis that the sleep effects of morphine interacting in distinct regions of the brainstem are opioid-receptor phenomena. Microdialysis studies performed in the medial pontine reticular formation have indicated that inhibition of acetylcholine release in this area could be a mechanism through which morphine may inhibit PS.  In addition to the well known EEG-synchronizing and SWS-enhancing effects mediated by the NST, [23,24] this region has a role in PS mechanisms, because both the administration of cholinergic agonists into the NST  and destruction of its serotoninergic innervation  produces enhancement of PS. Despite this and the fact that NTS contains muscarinic receptors,  this study shows that PS was unchanged after NST administration of the different opioids. No significant changes were detected in the amount of PS or the duration and number of its episodes. It seems that micro and delta opioids in the NST cause an enhancement of SWS that mimics the results observed after serotoninergic [24,25,47] and not acetylcholine stimulation.
Concerning the identification of the opioid receptors that mediated the increase in SWS caused by the NST morphine injections in cats, the present study showed that the microinjection of morphiceptin and DPDPE (highly specific agonists of micro- and delta-opioid receptors, respectively) into the NST provoked an increase in SWS like that produced by morphine. These effects were also dose-dependent and site-specific. However, microinjection of U-50488-H, the most selective kappa-opioid agonist,  into the NST did not produce any significant changes in the SWC states. It therefore seems that the SWS enhancement caused by NTS administration of morphine in cats is mediated by the micro- and delta-opioid receptors. However, further studies using specific antagonists to block the effects of morphiceptin and DPDPE are needed to provide unequivocal evidence of the involvement of both micro and delta receptors in the enhancement of SWS produced by opioids in the NST.
The NST has been reported to be basically enkephalinergic,  but the NST neurons can produce dynorphins and endorphins.  Nevertheless, the opioid-receptor subtype distribution varies among the different species. Thus, micro- and kappa-opioid receptors have been localized in the rat NST,  whereas micro and delta but not kappa receptors have been demonstrated in the NST of rabbits  and cats. [16,18] Therefore, the lack of effect of the specific kappa agonist U-50488-H in the current study would seem to be explained by the absence of kappa receptors in the feline NST.
The authors thank Professor F Reinoso-Suarez, for his support and advice; G. de la Fuente and M. T. Fernandez Yuste, for their assistance with the research and illustrations, respectively; and J. Hernandez-Claumarchirant, P. Romero, and R.M. Sanchez-Lozano, for their help with the histologic preparations. The U-50488-H was provided by Upjohn Laboratories.