Buprenorphine Disrupts Sleep and Decreases Adenosine Concentrations in Sleep-regulating Brain Regions of Sprague Dawley Rat

• Opiate-induced sleep disruption can contribute to hyperalgesia

• Endogenous adenosine promotes sleep and decreases nociception

• Antinociceptive doses of buprenorphine disrupted normal sleep architecture in rats, an effect that was attenuated by the sedative-hypnotic eszopiclone

• Buprenorphine also decreased adenosine levels in brain regions known to modulate sleep and nociception

OPIOIDS are used effectively in the treatment of chronic and acute pain, and the extensive use of opioids encourages efforts to develop countermeasures to combat unwanted side effects.1,2Opioids disrupt sleep,3,–,7and sleep disruption can contribute to hyperalgesia,8,–,16impaired immune function,17and postoperative cognitive impairment.18,19 

Adenosine is an endogenous neuromodulator that significantly enhances sleep20and diminishes nociception.21Sleep is increased by increasing adenosine in the pontine reticular formation (PnO)22,–,24and in the substantia innominata (SI) area of the basal forebrain.20,25Adenosine concentrations in the PnO and SI are decreased by the μ-opioid receptor agonists morphine and fentanyl.26 

Buprenorphine, a partial μ-opioid receptor agonist and κ-opioid receptor antagonist, is an effective analgesic, but no previous studies have quantified the effects of buprenorphine on sleep architecture7,27,28or on adenosine concentrations in the PnO and SI. This study was designed to test the hypothesis that buprenorphine decreases sleep and adenosine concentrations in PnO and SI, brain regions known to modulate sleep and nociception.

Adult, male Crl:CD(SD) (Sprague Dawley) rats (n = 26) purchased from Charles River Laboratories (Wilmington, MA) were used for all studies. Rats weighing 250–350 g were used because brains from rats in this weight range are known to fit the rat stereotaxic atlas.29Male rats were chosen to facilitate comparison of the current results to previous data obtained from males.26,30,–,32Rats were housed in a 12:12-h light–dark cycle (lights on from 8:00 to 20:00) with access to food and water ad libitum . Procedures were reviewed and approved by the University of Michigan Committee on the Use and Care of Animals. Every phase of this study adhered to the Guide for the Care and Use of Laboratory Animals  (Eighth Edition, National Academy of Sciences Press, Washington DC, 2011).‖

Rats were anesthetized with 3% isoflurane (Hospira, Inc., Lake Forest, IL). The jugular vein was exposed, and a catheter (12 cm of Micro-Renathane tubing [MRE–040], Braintree Scientific, Braintree, MA) was inserted in the direction of the heart. The other end of the catheter was tunneled subcutaneously and implanted between the scapulas. A back-mounted flange guide cannula (8I 1000BM10; Plastics One, Roanoke, VA) and dummy cannula (8IC313DCCACC; Plastics One) were secured with the catheter in the midscapular position. This procedure provided subsequent venous access.

Implantation of the jugular vein catheter was followed immediately by implantation of electrodes for recording sleep. Rats were moved to a Kopf Model 962 small animal stereotaxic instrument fitted with a Model 906 rat anesthesia mask (David Kopf Instruments, Tujunga, CA), and anesthesia was maintained with isoflurane (2.0%). Three electrodes (8IE36320SPCE, Plastics One) for recording cortical electroencephalogram were placed 2.0 mm posterior and 1.3 mm lateral to bregma, 2.0 mm posterior and 1.5 mm lateral to bregma, and 1.0 mm anterior and 1.5 mm lateral to bregma.29Two electrodes (4 cm of AG 7/40T Medwire, Mt. Vernon, NY) for electromyogram recordings were placed in the dorsal neck muscle, and a third electrode was placed under the skin of the neck muscle as a reference. The nonimplanted ends of the electroencephalogram and electromyogram electrodes were soldered to electrical contact pins (E363/0; Plastics One) that were plugged into a plastic pedestal (8K00022980IF; Plastics One). Three stainless steel anchor screws (MPX-0080–02P-C; Small Parts Inc., Miami Lakes, FL) were placed in the skull to secure the electrodes. Dental acrylic was used to construct a head cap covering the electrodes and to anchor the electrical connector and electrodes to the skull. Rats were then removed from the stereotaxic frame and monitored during recovery from anesthesia. Once ambulatory, animals were returned to their home cages.

Rats were given 1 week for surgical recovery and then conditioned for an additional week to 10 days to sleeping in a Raturn (Bioanalytical Systems, West Lafayette, IN) recording chamber. During conditioning, the implanted electrodes were attached by a cable (363–441/six 80CM 6TCM; Plastics One) to amplifiers and a computer for digital recording of electroencephalogram and electromyogram signals. Rats had free access to food and water while in the recording chambers.

An initial series of experiments was conducted to confirm that the 1 mg/kg dose of buprenorphine produced antinociception as reported previously.33Procedures for thermal nociceptive testing have been described in detail.34,35Briefly, rats were conditioned to being placed in the plexiglass chamber of a Hargreaves Paw Withdrawal unit (Model 336T; IITC Life Science, Woodland Hills, CA) 1 h each day for the week before data collection. The Model 400 (IITC Life Science) heated glass stand and base was set to 30°C for the last 10 min of each conditioning session, and both hind paws of the rat were exposed to the heat stimulus.36Five baseline measurements were taken after the habituation time. As soon as baseline measurements were recorded, saline or buprenorphine hydrochloride (Sigma–Aldrich, Saint Louis, MO; 1 mg/kg) was administered via  the jugular vein catheter. The injection volume was 1 ml. Measures of paw withdrawal latency (PWL) were taken at 10, 20, 30, 60, 90, and 120 min after saline or buprenorphine administration. A cutoff time of 15 s was set to prevent tissue damage of the hind paw.

A second series of experiments was designed to quantify the effect of intravenously administered buprenorphine on states of sleep and wakefulness. Buprenorphine was dissolved in sterile saline (pH 5.8 ± 0.2) and administered intravenously in a 1-ml volume at a dose of 1 mg/kg. Saline injection provided a negative control condition.

Recordings of sleep and wakefulness began at 08:00 at the initiation of the light phase of the light–dark cycle. Rats are nocturnal and light onset corresponds to the rat subjective night. To determine whether buprenorphine caused sleep disruption, as do other opioids,26,37this study was designed to deliver buprenorphine at light onset. Rats were placed in the recording chamber, and the electromyogram and electroencephalogram electrodes were attached via  swivel cable to the amplifiers and computer. All injections were administered during a 4-min interval. The data acquisition software was started when drug or vehicle administration began. The electroencephalogram signals were filtered between 0.3 and 30 Hz and amplified. Each rat (n = 7) received one injection of buprenorphine and one injection of saline separated by at least 1 week. The rats were then allowed to sleep and wake spontaneously for the remainder of the 24-h recording. At the end of the recording interval, rats were returned to the vivarium. Every 10 s of the 24-h recording was scored as wakefulness, nonrapid eye movement (NREM) sleep, or rapid eye movement (REM) sleep. All sleep recordings were also scored by a second individual who was blinded to the injection condition. There was 93% agreement between the two sleep scorers.

A third series of experiments was designed to coadminister the nonbenzodiazepine sedative-hypnotic eszopiclone (Toronto Research Chemicals, Toronto, Canada) with buprenorphine to quantify the effect on sleep and wakefulness. Eszopiclone is a benzodiazepine receptor agonist with a nonbenzodiazepine structure, marketed as Lunesta® (Sunovion Pharmaceuticals, Marlborough, MA). Eszopiclone is the (S)-isomer of the cyclopyrrolone zopiclone and is indicated for the treatment of insomnia.38As discussed in detail elsewhere,26patients who experience pain often report poor sleep. Clinically used doses of opioids significantly disrupt sleep,37and disordered sleep exacerbates pain.26,39,40These data raise the question of whether enhancement of sleep by a sedative-hypnotic would have a beneficial effect of diminishing opioid-induced sleep disruption. If so, this would encourage future studies aiming to determine whether combining opioid and sedative-hypnotic therapy could diminish pain. Eszopiclone was dissolved in sterile saline and dimethyl sulfoxide (1%; pH 6.0 ± 0.2) and administered intravenously (3 mg/kg). Buprenorphine (1 mg/kg) was then delivered via  the same IV cannula. For these studies, rats (n = 4) received an injection of eszopiclone followed immediately by an injection of saline or buprenorphine.

A fourth set of experiments sought to identify brain regions through which buprenorphine decreased sleep. Normal cholinergic neurotransmission is essential for maintaining wakefulness, and opioids disrupt cholinergic neurotransmission in the SI region of the basal forebrain.30Adenosine is known to promote sleep, and previous studies have shown that adenosine concentrations in the PnO are decreased by fentanyl and morphine.26Both fentanyl and morphine cause sleep disruption. Therefore, the current experiments also used in vivo  microdialysis and high-performance liquid chromatography to measure adenosine concentrations in the PnO and SI during dialysis delivery of buprenorphine.

Buprenorphine (100 μM) was prepared the morning of each experiment. The drug was dissolved in Ringer's solution (pH 5.8–6.2) comprised of 146 nM NaCl, 4.0 mM KCl, 2.4 mM CaCl², and 10 μM erythro-9-(2-hydroxy-3-nonyl)adenine (EHNA; Sigma–Aldrich), which is an adenosine deaminase inhibitor. Each rat was placed in an induction chamber and anesthetized with 4% isoflurane (Hospira, Inc.) in 100% oxygen. After 5 min, the rat was moved from the chamber into a stereotaxic frame and fitted with a rat anesthesia mask, as described. The isoflurane concentration was reduced to 2.5%. A midline scalp incision was then made to expose lambda and bregma. A rotary tool (Dremel, Racine, WI) was used to make a small craniotomy, through which a dialysis probe was placed in the brain. A rat brain atlas29was used to position a CMA-11 microdialysis probe (Cuprophane membrane: 1 mm long, 0.24 mm in diameter, 6-kDa cutoff; CMA Microdialysis, North Chelmsford, MA) in the PnO or in the SI. Stereotaxic coordinates for the PnO were 8.4 mm posterior to bregma, 1.0 mm lateral to the midline, and 9.2 mm below bregma. The coordinates for the SI were 1.6 mm posterior to bregma, 2.5 mm lateral to the midline, and 8.7 mm below bregma. The delivered isoflurane concentration was held at 1.5% and measured continuously throughout the duration of the experiment. A water blanket and recirculating heat pump (Gaymar Industries, Orchard Park, NY) were used to maintain body temperature at 37°C throughout data collection and recovery.

The dialysis probe was perfused with Ringer's solution at a constant flow rate of 2 μl/min using a CMA/400 pump. Dialysis intervals of 15 min produced 30-μl samples, which were injected into an high-performance liquid chromatography system coupled to a UV-Vis detector (wavelength of 254 nm). This system made it possible to express measured adenosine concentrations as nanomolar (nM). The digitized chromatographs were quantified against a standard curve using ChromGraph software (Bioanalytical Systems). Adenosine concentrations were allowed to stabilize for 2 h before beginning data collection. A control sample was collected every 15 min for 1 h during dialysis with Ringer's solution. At the end of the fourth control sample, a liquid switch was activated to begin dialysis with Ringer's solution containing buprenorphine (100 μM). As noted elsewhere,26,41,42the characteristics of the dialysis membrane are such that only approximately 5% of the 100 μM buprenorphine was delivered to the brain. After the final dialysis sample was collected, the probe was removed from the brain and the scalp incision was closed. The delivery of isoflurane was discontinued, and the animal was removed from the stereotaxic frame. The rats were returned to their cages and monitored until they were ambulatory.

Four to 5 days after the microdialysis experiment, each rat was deeply anesthetized and decapitated. Brains were removed, cut into 40-μm thick coronal sections with a cryostat (Leica Microsystems, Nussloch, Germany), and slide-mounted serially. Slides with brain sections containing the SI and PnO were fixed in paraformaldehyde vapor (80°C) and stained with cresyl violet. All sections were then digitized using a Nikon Super Coolscan 4000 scanner (Tokyo, Japan). Coronal sections were compared with plates in a rat brain atlas,29and the dialysis sites were localized to either the PnO or the SI.

Statistical programs used for data analysis included Prism 5 (Graph Pad Software, Inc., La Jolla, CA) and SAS v9.2 (SAS Institute Inc., Cary, NC). All data were tested to ensure they met the assumptions of the underlying statistical model. PWL was converted to percent maximum possible effect (% MPE) using the following equation: % MPE = (PWL experimental − PWL baseline)/(15 s − PWL baseline) × 100. Repeated measures, two-way analysis of variance (ANOVA) was used to analyze results for changes over time and changes caused by the drug, and Bonferroni post hoc  tests were used to detect differences at specific time points.

Every 10 s of the 24-h sleep and wakefulness recording was scored as wakefulness, NREM sleep, or REM sleep. Dependent measures included percent of time spent in each state, latency to onset of the first episode of NREM sleep and REM sleep, number of episodes, average episode duration, and number of transitions between states. To avoid the problem of inflated degrees of freedom resulting from the large number of 10-s epochs analyzed, the sleep–wake data were averaged for each rat. Dependent measures of sleep and wakefulness were analyzed by repeated measures two-way ANOVA and paired t  tests using Bonferroni correction.

As described in detail elsewhere,24,43,44fast Fourier transform of the electroencephalogram was performed to determine whether the electroencephalogram was altered by buprenorphine. Electroencephalographic power was analyzed by repeated measures two-way ANOVA and post hoc  tests for comparison at every 0.5-Hz frequency band (wakefulness and REM sleep 5.0–10.0 Hz; NREM sleep 0.5–5.0 Hz).

For each experiment, adenosine measures during dialysis with Ringer's solution (control) were compared with adenosine concentrations during dialysis delivery of buprenorphine. This design allowed each experiment to contribute one mean adenosine value derived from four control (Ringer's solution) samples and one mean adenosine value derived from four measures obtained during administration of buprenorphine. These values were then averaged across multiple experiments and analyzed individually for PnO and SI brain regions using paired t  tests. A probability value of P < 0.05 was considered statistically significant.

Figure 1depicts % MPE for PWL as a function of time after IV administration of saline and buprenorphine. ANOVA revealed that buprenorphine caused significant (P = 0.0072) antinociception. Bonferroni post hoc  comparisons indicated that buprenorphine significantly (P < 0.05) increased % MPE at 20, 30, 60, and 120 min after injection. This antinociceptive dose of buprenorphine was used for subsequent studies of sleep and wakefulness.

Figure 2illustrates the temporal distribution of wakefulness, NREM sleep, and REM sleep for 24 h after IV administration of saline (control) and buprenorphine. Figure 3summarizes group data for the light phase (first 12 h after injection) showing buprenorphine-induced changes in the temporal organization of sleep and wakefulness. ANOVA indicated a significant (P < 0.01) effect of buprenorphine on percent of time spent in states of wakefulness, NREM sleep, and REM sleep, as well as a significant (P < 0.0001) drug-by-state interaction (fig. 3A). Paired t  tests with Bonferroni correction showed that buprenorphine significantly (P < 0.001) increased the percentage of time spent in waking (25.2%) and significantly decreased the amount of time spent in NREM sleep (−22.1%) and REM sleep (−3.1%). Buprenorphine significantly delayed the onset of NREM sleep and REM sleep (fig. 3C).

There was a significant (P < 0.0001) drug main-effect and state-by-drug interaction (P < 0.0001) for the number of sleep–wake episodes (fig. 3E). Buprenorphine decreased the number of episodes of wakefulness (−88.2%), NREM sleep (−89.5%), and REM sleep (−90.8%). Figure 3G shows that buprenorphine significantly (P < 0.0001) altered the duration of sleep–wake episodes. The average duration of wakefulness was significantly increased (529.6%), and the duration of sleep epochs was decreased for both NREM sleep (−30.8%) and REM sleep (−87.5%). Figure 3I shows that buprenorphine also significantly (P < 0.0001) decreased the number of transitions (−89.8%) between states.

Figure 4plots the percentage state for each drug condition during the 12-h dark phase (rat subjective day) of the light–dark cycle that followed the 12-h light phase depicted by figure 3. Within the dark phase, when rats are normally awake and active, the time spent awake was significantly (P = 0.012) decreased by buprenorphine. The buprenorphine condition within the dark phase revealed significantly (P = 0.0019) more NREM sleep and a nonsignificant decrease in REM sleep compared with the saline condition.

The effect of buprenorphine on states of sleep and wakefulness can also be visualized by comparing the light phase (fig. 3A) and dark phase (fig. 4) results. NREM sleep after buprenorphine increased significantly (P = 0.0003) from an average of 5.5% in the light phase (fig. 3A) to 27.4% in the dark phase (fig. 4). There was also a significant (P = 0.003) rebound increase in REM sleep, from an average of 0.33% after buprenorphine during the light phase (fig. 3A) to approximately 4% after buprenorphine during the dark phase (fig. 4).

The five illustrations in the right column of figure 3summarize the results of experiments designed to determine whether the sedative-hypnotic eszopiclone countered the buprenorphine-induced inhibition of sleep. Eszopiclone when coadministered with buprenorphine prevented the significant increase in wakefulness (fig. 3B) caused by buprenorphine alone (fig. 3A). Similarly, the significant buprenorphine-induced decrease in NREM sleep and REM sleep (fig. 3A) was prevented by coadministration of eszopiclone (fig. 3B). Eszopiclone blocked the significant increase in latency to sleep onset (compare fig. 3C and D). Eszopiclone partially reversed the buprenorphine-induced decrease in the number of wakefulness and NREM sleep episodes (compare fig. 3E and F). The 530% increase in average duration of waking episodes caused by buprenorphine (fig. 3G) was reduced to a 171% increase by coadministration of eszopiclone (fig. 3H). Eszopiclone blocked the significant decrease in number of state transitions caused by buprenorphine (compare fig. 3I and J).

Figure 5A–C illustrates electroencephalogram power recorded across states of sleep and wakefulness after IV administration of buprenorphine to awake, freely moving rats. Buprenorphine did not alter electroencephalogram power during wakefulness or REM sleep (fig. 5, A and C) but did increase electroencephalogram power in the delta frequency range during NREM sleep (fig. 5B). ANOVA revealed a significant (P = 0.007) buprenorphine main effect on electroencephalogram frequency bands ranging from 0.5 to 5.0 Hz in 0.5-Hz increments (fig. 5B). The fast Fourier transform analyses were conducted for electroencephalogram measures obtained during the 12-h light period (i.e. , rat's subjective night) that immediately followed buprenorphine administration. As figures 2and 3show, buprenorphine depressed NREM sleep for 6–8 h. Measurement of the increase in electroencephalogram delta power was conducted for as long as 12-h after buprenorphine administration. A future study will be needed to determine whether, and for how long beyond 12-h, electroencephalogram delta power is increased by buprenorphine.

Histologic analyses confirmed that all microdialysis sites were localized to the PnO or SI (fig. 6A). Figure 6B shows the results of one representative experiment. Adenosine concentrations in the SI are plotted as a function of time during dialysis with Ringer's solution (121–180 min after probe placement) followed by dialysis delivery of buprenorphine (181–240 min after probe placement). Figure 6, C and D, confirms chromatographic identification of adenosine. Figure 6C illustrates chromatograms produced by five known concentrations of adenosine. Figure 6D shows chromatograms reflecting brain adenosine (dialyzed Ringer's solution), a negative control (nondialyzed Ringer's solution), a positive control (brain adenosine sampled during dialysis delivery of the adenosine deaminase inhibitor EHNA), and an adenosine standard.

Figure 7A summarizes a final set of experiments that quantified adenosine concentrations in SI and PnO as a function of route of buprenorphine administration. Microdialysis delivery of buprenorphine significantly (P = 0.03) decreased adenosine concentrations in PnO (−14.8%) and significantly (P = 0.0004) decreased adenosine concentrations in the SI region of the basal forebrain (−36.7%). Figure 7B plots adenosine concentrations in the SI before and after IV administration of buprenorphine to isoflurane-anesthetized rat. Buprenorphine significantly (P < 0.0001) decreased (−20.3%) adenosine concentrations in the SI.

Buprenorphine can be efficacious in the treatment of opioid and heroin addiction,45,–,47and there is increasing interest in the use of buprenorphine for pain management.27,28The analgesic effects of buprenorphine are mediated, in part, via  agonist actions at the μ-opioid receptor.48This is the first study presenting electrographic data that demonstrate significant sleep disturbance (fig. 3) caused by an antinociceptive dose of buprenorphine (fig. 1). The current finding that an antinociceptive dose of buprenorphine disrupts sleep is discussed relative to the relationship between sleep and nociception, the potential for developing countermeasures for opioid-induced sleep disruption, and the underlying mechanisms.

Some data suggest that buprenorphine is superior to traditional opioids for the treatment of pain because of its reported analgesic and antihyperalgesic effects with fewer side effects (low incidence of respiratory depression and less constipation).28Buprenorphine shares some similar pharmacodynamic properties with traditional opioids, and patient-report data indicate benefits from buprenorphine therapy. Freye et al.  27found that self-report sleep quality rated as “good” or “very good” increased from 14% to 74% when patient regimens were transitioned from high-dose oral morphine to transdermal buprenorphine. Transdermal buprenorphine has been compared with placebo for ability to decrease pain and promote sleep, and patients randomized to receive buprenorphine report less pain and improved sleep.49Specifically, subjects who received buprenorphine reported less trouble falling asleep, decreased requirement for sleeping medication, and decreased awakening at night because of pain. Another study found a nonsignificant trend of improved sleep favoring transdermal buprenorphine over extended-release tramadol tablets for the treatment of osteoarthritis.50A known limitation of such studies is that self-assessment of sleep quality may not show faithful concordance with objective, electrographic measures of sleep.51 

There is a growing appreciation for the interrelationship between sleep and pain.39Sleep deprivation in healthy normal individuals decreases pain perception thresholds.52The chronic effects of μ-opioid receptor agonists on sleep in patients with pain are not understood completely.39Opioids cause sleep disturbance4,37,53and the current results demonstrate that buprenorphine increases wakefulness and disrupts the temporal organization of sleep (figs. 2–4). Sleep, like breathing, is an endogenously generated biologic rhythm. Just as rhythmic switching from inspiration to expiration is essential for gas exchange, the ability of sleep to produce reports of rest and recovery requires a normal temporal organization. Buprenorphine caused a decrease in the number (fig. 3E) and an increase in the duration (fig. 3G) of wakefulness episodes. The decreased number of state transitions (fig. 3I) reflects the buprenorphine-induced disruption of sleep continuity. Figure 4summarizes the percentage of time spent in states of sleep and wakefulness during the 12-h dark phase when rats are normally active. These dark-phase recordings were continuous with the figure 3data during the 12-h light phase. Thus, the figure 4data show that for 12–24 h after administration of buprenorphine, there was a rebound increase in sleep at a time when nocturnal rodents are normally most active. The potential clinical relevance of buprenorphine-induced disruption of sleep continuity derives from repeated sleep disruption negatively affecting neurocognitive function as severely as does total sleep deprivation.54 

Opioid-induced sleep disruption has the potential to negatively affect patient care because sleep deprivation is known to decrease pain threshold.15,16,52This study did not address the impact of pain or the treatment of pain on sleep disturbance. Some believe that medications from the agonist-antagonist class, such as buprenorphine, may be less associated with the adverse effects of traditional μ-agonists; however, the current data indicate that the sleep-disrupting effects of buprenorphine are similar to those of other opioids.26,30,37,53The current results encourage studies designed to objectively quantify the effects of buprenorphine on sleep in humans.

The U.S. Food and Drug Administration approved buprenorphine for the treatment of opioid addiction. Suboxone® (buprenorphine-naloxone sublingual tablet, Reckitt Benckiser Pharmaceuticals Inc., Richmond, VA), Subutex® (buprenorphine sublingual tablet, Reckitt Benckiser Pharmaceuticals Inc.), and transdermal buprenorphine (not available in the United States) are increasingly used to treat pain. The current finding of buprenorphine-induced sleep disruption also is relevant to evidence indicating that sleep disturbance can lead to a higher potential for addiction relapse.55,–,57 

Electroencephalogram delta waves (0.5–4 Hz) are one of three rhythmic waveforms characteristic of NREM sleep.58Delta waves provide an index of sleep intensity,59,60and electroencephalogram power in the delta frequency increases during recovery sleep that follows sleep deprivation.61,62The finding that buprenorphine caused an increase in electroencephalogram delta power during NREM sleep (fig. 5B) is consistent with subjective reports that buprenorphine improves sleep7,27and with evidence that opioids increase delta power in humans.63A search of the Medline database from 2001 to 2011 revealed no polysomnographic data characterizing the effect of buprenorphine on human sleep.

Buprenorphine significantly disrupted the temporal organization of sleep (fig. 3, panels A, C, E, G, and I). Most aspects of sleep disturbance were prevented by concomitant treatment with the nonbenzodiazepine hypnotic eszopiclone (fig. 3, panels B, D, F, H, and J). Sedative-hypnotics are a standard treatment for insomnia, but their effects in the treatment of pain- and opioid-induced sleep disturbance remain poorly understood. The use of sedative-hypnotics may not be a mainstay of addiction therapy, but the current results (fig. 3, panels B, D, F, H, and J) indicate their potential to prevent buprenorphine-induced sleep disturbance. Eszopiclone as an adjunctive agent coadministered with the antidepressant fluoxetine resulted in a faster onset and greater magnitude of the desired antidepressant effect.64An exciting area open to future study is to determine whether hypnotics can be used as an effective countermeasure for opioid-induced sleep disruption.

The current study was designed to quantify the effects of buprenorphine on sleep and adenosine concentrations. The results are limited to documenting that buprenorphine, similar to morphine and fentanyl, disrupted sleep and decreased adenosine concentrations in sleep-related brain regions. The results do not imply that the effects of buprenorphine were mediated only by μ-opioid receptors. Buprenorphine may have disrupted sleep and decreased adenosine, in part, by acting as a κ antagonist.

The analgesic21and sleep-promoting65effects of adenosine are well known and suggest adenosine as a molecule of potential clinical relevance for anesthesiology. There is good agreement between preclinical and clinical data that opioids disrupt sleep,37a finding confirmed by administering opioids to pain-free humans.53The restorative effects of sleep require normal temporal organization of sleep. Unfortunately, morphine and fentanyl slow the electroencephalogram during wakefulness, increase lighter stage 2 NREM sleep, decrease stage 3 and 4 NREM sleep, and decrease or eliminate REM sleep. Disruption of normal sleep impairs immune function, exacerbates pain,39and, particularly in older patients, can be a precipitating factor for postoperative delirium.66 

In conclusion, the results show for the first time that buprenorphine disrupted normal sleep architecture and decreased adenosine concentrations in sleep-regulating regions of the basal forebrain and PnO (figs. 6and 7). The buprenorphine results are consistent with the discovery that fentanyl and morphine decrease adenosine concentrations in basal forebrain and PnO.26The current study extends the previous findings by providing mechanistic insights into a brain site and a molecule by which buprenorphine disrupts sleep. Novel insights were obtained by holding the site of adenosine measurement constant within the SI region of the basal forebrain while varying the route of buprenorphine delivery. The results show that both microdialysis delivery to the SI and systemic administration of buprenorphine caused a significant decrease in adenosine in the SI. As demonstrated elsewhere,30,44when effects caused by drug delivery to a specific brain region replicate the effects caused by systemic delivery, it is logical to infer that the actions of systemically administered drugs are mediated, in part, by that brain region and by the neurotransmitter molecule being measured. Thus, the neurochemical results, combined with the sleep-disrupting effect of buprenorphine, support the interpretation that one mechanism by which buprenorphine disrupts sleep is decreasing adenosine concentrations in the SI region of the basal forebrain.

For expert assistance, the authors thank Sha Jiang, B.S., Research Associate, Mary A. Norat, B.S., Senior Research Associate, and Sarah L. Watson, B.S., Senior Research Associate, Department of Anesthesiology; and Kathy Welch, M.A., M.P.H., Statistician Staff Specialist, Center for Statistical Consultation and Research, University of Michigan, Ann Arbor, Michigan.