Prostaglandins (PG) are members of the eicosanoid family derived mainly from arachidonic acid by the action of cyclooxygenase (COX), which produces prostaglandin H2 (PGH2), which is then transformed by a range of specific enzymes into the primary prostanoids such as prostaglandin D2 (PGD2) and E2 (PGE2). There are two isoforms of COX: COX-1 is constitutive and predominant in most tissues whereas COX-2 is inducible and tightly regulated at transcription/translation levels.
Ethanol-induced loss of righting reflex (LORR) in mice was significantly decreased by COX inhibitors or nonsteroidal antiinflammatory drugs (NSAIDs) such as aspirin, indomethacin, mefenamic acid, and flufenamic acid (1). In addition, prostanoids could enhance the sedative effects of ethanol (2). Endogenous PG levels were increased by ethanol to a larger extent in LS mice (“long sleep” mice sensitive to ethanol) than in SS mice (“short sleep” mice insensitive to ethanol) (3). Therefore, these observations indicate that ethanol, at least in part, mediates its central nervous system (CNS) depressant effects by increasing the production of PGs. However, similar results were not obtained for other anesthetic drugs such as pentobarbital and chloral hydrate (1).
The benzodiazepines are the most widely used CNS depressant drugs. They act specifically on the benzodiazepine receptor of the gamma-aminobutyric acid (GABA)A receptor complex and thus modify the GABA-induced chloride flux. However, it has been demonstrated that muscimol, a selective GABAA receptor agonist, stimulates glioma C6 cells to release arachidonic acid, which is converted to PGD2 in the presence of diazepam (4). PGD2 is an endogenous sleep promoting substance (5). It is interesting, therefore, to investigate whether PGs play a role in the CNS depressant effects of diazepam. We report here that COX-1 inhibitors decreased whereas arachidonic acid increased diazepam-induced LORR in mice.
This work was approved by the Animal Ethics Committee of National University of Singapore and all efforts were made to minimize the number of mice used. Male Swiss-Albino mice (25 to 30 g) were obtained from the Laboratory Animal Center and housed in groups of 5 with food and water available ad libitum on a 12-h light-dark cycle. Mice were assigned to experimental groups at random and each was used in only one experiment.
All drugs were administered by intraperitoneal (IP) injection with an injection volume of 5 mL/kg. Aspirin, arachidonic acid and small-dose diazepam were dissolved in 30% dimethyl sulfoxide (DMSO). All other drugs were dissolved in DMSO. Control mice received the appropriate vehicle only. Diazepam was administered 30 min after pretreatment.
After the administration of diazepam (20 mg/kg), mice were put on their backs to test for LORR. The righting reflex was considered lost if the mice could not right themselves within 1 min, and regained when they managed to right themselves 3 times within 1 min. The duration of LORR was recorded for each mouse. All experiments were conducted between 1000 and 1200 h.
Mice were screened 24 h before experimentation for their ability to stay on a rotarod (Acceler Rota-rod, Ugo Basile, Comerio, Italy), which accelerated from 2 to 20 revolutions/min in 5 min. Only mice that could stay on for the full 5 min were used in the experiment. After the administration of diazepam (5 mg/kg), mice were put into an activity cage. Spontaneous ambulation was monitored for 3 min after a delay of 1 min. The activity cage (Columbus Instruments, Columbus, OH) consisted of a transparent polypropylene box of dimension 40 (l) × 20 (w) × 20 (h) cm. The movement of a mouse inside the box was monitored by 15 infrared beams placed equidistant, and the distance between the first and the last beam was 38 cm. Beam interceptions were recorded automatically by a computer-based system. Ambulatory activities were recorded at 30-s time blocks. Three min after the completion of the activity cage measurements, the mouse was then placed onto the rotarod and the fall latency was recorded; otherwise, the mouse was given a latency of 5 min. All experiments were conducted between 1900 and 2300 h.
Mice were given an IP injection (4 mL/kg) of either saline (control) or diazepam (20 mg/kg). They were killed by cervical dislocation 30 min after the injection. Cerebral cortex and hypothalamus/thalamus were collected and homogenized (Heidolph DIAX 900 Homogenizer, Heidolph Instruments LLC, Cinnaminson, NJ) in ice-cold phosphate buffer containing 10 μM indomethacin. The homogenates were then centrifuged at 15,000 rpm and the supernatant was removed and stored at −20°C until the PG assay. Indomethacin inhibits both COX enzymes and 15-hydroxyprostaglandin dehydrogenase that catalyzes the catabolism of PG and was added to minimize errors resulting from the possible synthesis and degradation of PG in the homogenate. The concentration of PGD2 was determined using commercial kits according to manufacturer protocol (Cayman Chemical, Ann Arbor, MI). The PGE2 kits commercially available use anti-mouse IgG. To minimize possible cross-reaction, a modified PGE2 methoxime enzyme immunoassay kit was used with reagents purchased from Cayman.
Mice were treated as described in the previous section. Total RNA was extracted from the cerebral cortex using Trizol reagent. RNA concentration and purity were assessed by measuring the absorbance at 260 and 280 nm wavelength (Shimadzu UV-1601 UV-visible Spectrophotometer Shimadzu Corp., Tokyo, Japan). Only RNA with absorbance ratio between 1.8 and 2.0 was used for further reverse transcriptase-polymerase chain reaction (RT-PCR) analysis.
Aliquots of total RNA (5 μg) were used in the RT reaction mixed with 0.5 μg oligo(dT) primer. First-strand cDNA synthesis was then performed with the use of Avian Myeloblastosis Virus RT (Invitrogen, Carlsbad, CA). cDNA was then amplified by PCR using primers specific to the key enzymes in PG synthesis pathway. β-actin, a ubiquitously expressed gene, was used as an internal control. The specificity of each pair of primers was tested by blasting their sequences to the entire mice genome. The quality of the primers was tested by the software, NetPrimer (Shimadzu).
β-actin primers were as follows: forward, 5′-ATCTGGCACCACACCTTCTACAATGAGCTGCG-3′; reverse: 5′-TACTCCTGCTTGCTGATCCACATCTGC-3′, which yield a PCR product of 832bp. Cox-1 primers were as follows: forward, 5′-AGGAGATGGCTGCTGAGTTGG-3′; reverse: 5′-AATCTGACTTTCTGAGTTGCC-3′, which result in a PCR product of 601bp.
Lipocalin type-PGD synthase (L-PGDS) primers were as follows: forward, 5′-CAGGAAAAACCAGTGTGAGACC-3′; reverse, 5′-AGAGGGTGGCCATGCGGAAG-3′, which yield a PCR product of 194bp. Membrane-associated PGE synthase-1 (mPGES-1) primers were as follows: forward, 5′-TGTACGCGGTGGCTGTCATC-3′; reverse, 5′-GCCAGGACATAGGCCCCGG-3′, which result in a PCR product of 319bp. 2 μL from each RT reaction mixture was mixed with 1 μL of each primer (20 mM), 25 μL PCR reaction mix and the final volume was topped up to 50 μL. To remain in the linear range during amplification, the mixture was heated at 95°C for 5 min and the reaction was performed respectively as follows: 29 cycles (95°C for 30 s; 60°C for 30 secs and 72°C for 1 min; 10 min for the last extension) for β-actin; 30 cycles (95°C for 30 s; 55°C for 30 s; and 72°C for 1 min; 10 min for the last extension) for COX-1, 29 cycles (95°C for 30 s; 60°C for 30 s; and 72°C for 1 min; 10 min for the last extension) for L-PGDS and 39 cycles (95°C for 30 s; 60°C for 30 s; and 72°C for 1 min; 10 min for the last extension) for mPGES-1. The amplified PCR products were then separated by electrophoresis (Bio-Rad laboratories, Hercules, CA) on 1% agarose gel. A Gene Ruler 1000-bp DNA Ladder was used to determine the size of the PCR products. Densitometry of the bands was performed (Syngene Multi Genius Bioimaging System, Frederick, MD) to give rough quantifications. The expression level of each gene was estimated as its ratio to β-actin. Each gene in each sample was amplified twice by PCR and the average of the two ratios was taken and used for statistical analysis. To minimize possible influence of natural circadian cycle to the gene expression, all individual experiments were started at between 1400 and 1500 h.
Data are expressed as mean ± sd. All statistical analyses were performed by SPSS for Windows, Version 12 (SPSS, Chicago, IL). One-way analysis of variance or 2-way analysis of variance with repeated measures was followed by post hoc analysis with Bonferroni correction. In all cases, P < 0.05 was taken as the level of significance.
Diazepam (20 mg/kg) caused a LORR for 120 ± 12 min in control mice pretreated with vehicle (30% DMSO). Pretreatment with aspirin (10 mg/kg but not 2 mg/kg) significantly reduced the duration of LORR by 31% to 83 ± 10 min (P < 0.002) (Fig. 1). Pretreatment with arachidonic acid (20 mg/kg), however, significantly increased the duration of LORR by 48% from 111 ± 19 min to 164 ± 35 min (P < 0.001) (Fig. 2). As expected, coadministration of aspirin abolished the effect of arachidonic acid.
Other COX inhibitors were administered in DMSO due to their insolublility in aqueous solvents. Control mice pretreated with DMSO showed markedly enhanced diazepam-induced LORR (275 ± 48 min or 2.3-fold) compared with mice pretreated with 30% DMSO (Figs. 1 and 2). This is consistent with the finding that DMSO enhanced the effects of barbiturates (6). However, despite the differences in control value, ibuprofen, indomethacin, and picroxicam at the same dose of 10 mg/kg, gave similar results by reducing the diazepam-induced LORR by 46%, 40%, and 41%, respectively (P < 0.005, Fig. 3).
Valeryl salicylate, a selective COX-1 inhibitor at 10 mg/kg, also reduced diazepam-induced LORR by 29% to 195 ± 14 min (P < 0.05). In contrast, NS398, a selective COX-2 inhibitor, failed to produce any effects even when the dose was doubled to 20 mg/kg (Fig. 3).
Spontaneous ambulatory activity was significantly decreased by diazepam (5 mg/kg) over the entire observation period of 3 min (P < 0.05). Pretreatment with aspirin did not significantly alter the effect of diazepam on spontaneous ambulation (P > 0.7) (Fig. 4A).
Consistent with reduced spontaneous motor activity, diazepam-injected mice showed significantly impaired performance on the rotarod, with a 25% decrease in the fall latency (P < 0.05). Again, aspirin pretreatment did not significantly alter the diazepam effect on the rotarod performance (Fig. 4B).
The cerebral cortex and hypothalamus/thalamus are responsible for the regulation of consciousness and sleep/wake cycle, respectively (7). They are thus selected for the measurement of PGD2 and PGE2 and the expression of related enzymes. Endogenous PGD2 and PGE2 levels in the cerebral cortex and hypothalamus/thalamus homogenates were determined at 30 min after injection. PGD2 was found to be 31% less (P < 0.05) in the cortex but not in the hypothalamus/thalamus. In contrast, PGE2 remained unchanged in both brain regions (Fig. 5). RT-PCR analysis revealed that the expression of COX-1, L-PGDS and m-PGES-1 were not affected by diazepam (Fig. 6).
In preliminary dose-finding experiments, it was found that diazepam at 20 mg/kg gave consistent mean duration of LORR (90–120 min). This is appropriate for studying a potential reduction of effect by COX inhibitors. With smaller doses, there were a significant number of mice that failed to show LORR. This would cause large variations in the duration of LORR making statistical comparison between groups difficult. The human equivalent dose (HED) of 20 mg/kg in mice is 1.62 mg/kg calculated on the basis of body surface area, or approximately 113 mg for a 70-kg person. This is thus a very large dose compared with the usual clinical doses (2–10 mg) being used in patients. However, this is consistent with the reported 95% effective dose of propofol at 200 mg/kg (HED 16.2 mg/kg) for inducing LORR in mice (8), whereas its clinical dose range is 1–2.5 mg/kg in humans. At 5 mg/kg, diazepam gave a suitable reduction in locomotion and rotarod performance for the purpose of this study. Dosages of 1–10 mg/kg are commonly used as motor impairing or sedative doses in mice (9).
It is well established that the benzodiazepines cause CNS depression by acting on the benzodiazepine receptors that are regulatory sites on the GABAA receptors. When diazepam binds to this site, it enhances GABA action on this receptor, leading to increased opening frequency of the receptor chloride channel and thus hyperpolarization of the neuronal membrane. The effects of the benzodiazepines are thus limited by the action of endogenous GABA, and they are not known to produce nonspecific effects related to its lipid solubility even at very large doses (10). It is therefore intriguing that arachidonic acid could enhance, while COX inhibitors could attenuate, the hypnotic effect of diazepam. This is consistent with previous findings that PG synthesis inhibitors attenuated ethanol-induced LORR (11) whereas PGD2 enhanced pentobarbital-induced LORR (12).
Although these findings clearly suggest that the arachidonic acid cascade can modulate the hypnotic effect of diazepam, the mechanism is unclear. Our present findings strongly suggest that one or more COX products, most likely PG, play a significant role in modulating the duration of LORR. The fact that the selective COX-2 inhibitor NS398 was ineffective, while the COX-1 selective inhibitor, valeryl salicylate, and all the nonselective inhibitors tested were effective, demonstrated the exclusive involvement of the constitutive COX-1 and not the inducible COX-2 enzyme. It is also interesting to note that aspirin did not alter the effect of small-dose diazepam on spontaneous ambulation and rotarod performance. This argues against any direct antagonistic action between aspirin and diazepam because the same aspirin dose attenuated the LORR induced by diazepam at a 20 mg/kg dose but not the spontaneous activity and rotarod performance at a 5 mg/kg dose.
PGD2 and PGE2 are major PGs in the mammalian brain (13,14). PGD2 was first reported to be a sleep-inducing substance in the rat brain (15) and then, subsequently, in the monkey (16), rabbit (17), and cat (5) brains. Infusion of PGD2 into the preoptic area of the hypothalamus induced sleep during the day or night, which is indistinguishable from natural sleep (5,18). PGD2 produced by brain-specific L-PGDS circulates in the cerebrospinal fluid and exhibits circadian fluctuation in parallel with the sleep-wake cycle (19). In this connection, there have been reports of decreased or disrupted sleep after administration of NSAIDs in both humans and animals during the normal sleep period (20,21).
Conversely, infusion of PGE2 into the cerebroventricle (22) increased wakefulness. Moreover, AH6809 (a PGE2 antagonist) increased sleep in rats (23) and PGE2 inhibited sleep in narcoleptic dogs (24). At the molecular level, PGE2 may cause wakefulness by activating arousal neurons in the tuberomamillary nucleus located at posterior hypothalamus via AMPA-type excitatory amino acid receptors (25). Therefore, it appears that these two PGs with opposing effects on sleep may be intimately involved in the regulation of the sleep-wake cycle (26).
Based on the known actions of PGD2 and PGE2, they are two likely candidates among COX-1 products that could be involved in the modulation of diazepam-induced LORR, perhaps an increase in PGD2 and/or a decrease in PGE2 levels. On the contrary, we observed only a small but significant decrease in PGD2 levels in the cerebral cortex 30 min after the administration of diazepam. The exact mechanism by which diazepam interacts with the arachidonic acid cascade remains to be elucidated. This would further enhance our understanding of the pharmacology of benzodiazepines.
The authors are grateful to W. L. Ting for her excellent technical assistance.
1. George FR, Collins AC. Prostaglandin synthetase inhibitors antagonize the depressant effects of ethanol. Pharmacol Biochem Behav 1978;10:865–9.
2. Segarnick DJ, Cordasco DM, Rotrosen J. Prostanoid modulation (mediation?) of certain behavioral effects of ethanol. Pharmacol Biochem Behav 1985;23:71–5.
3. George FR, Collins AC. Ethanol's behavioural effects may be partly due to increases in brain prostaglandin production. Alcohol Clin Exp Res 1985;9:143–6.
4. Majewska MD, Chuang DM. Benzodiazepines enhance the muscimol-dependent activation of phospholipase A2
in glioma C6
cells. J Pharmacol Exp Ther 1985;232:650–5.
5. Hayaishi O. Molecular mechanisms of sleep-wake regulation: roles of prostaglandins D2
. FASEB J 1991;5:2575–81.
6. Shan HC, Lal H. Effects of 1,1,1-trichloroethane administered by different routes and in different solvents on barbiturate hypnosis and metabolism in mice. J Toxicol Environ Health 1976;1:807–16.
7. McCormick DA, Bal T. Sleep and arousal: thalamocortical mechanisms. Ann Rev Neurosci 1997;20:185–215.
8. Irifune M, Takarada T, Shimizu Y, et al. Propofol-induced anesthesia in mice is mediated by gamma-aminobutyric acid-A and excitatory amino acid receptors. Anesth Analg 2003;97:424–9.
9. Kralic JE, O'Buckley TK, Khisti RT, et al. GABAA
receptor alpha-1 subunit deletion alters receptor subtype assembly, pharmacological and behavioral responses to benzodiazepines and zolpidem. Neuropharmacology 2002;43:685–94.
10. Haefely WE. Pharmacology of the benzodiazepine receptor. Eur Arch Psychiat Neurol Sci 1989;238:294–301.
11. Suzuki T, Shiozaki Y, Misawa M. Relationships between ethanol-induced sleep and aspirin, indomethacin or PGE2
in inbred rats. Arukoru Kenkyuto Yakubutsu Ison 1989;24:480–9.
12. Hollingsworth EB, Patrick GA. Involvement of the serotonergic system in the prolongation of pentobarbital sleeping time produced by prostaglandin D2
. Pharmacol Biochem Behav 1985;22:365–70.
13. Abdel-Halim MS, Hamberg M, Sjoquist B, Anggard E. Identification of prostaglandin D2
as a major prostaglandin in homogenates of rat brain. Prostaglandins 1977;14:633–43.
14. Ogorochi T, Narumiya S, Mizuno N, et al. Regional distribution of prostaglandins D2
, and F2α
and related enzymes in postmortem human brain. J Neurochem 1984;43:71–82.
15. Ueno R, Honda K, Inoue S, Hayaishi O. Prostaglandin D2
, a cerebral sleep-inducing substance in rats. Proc Natl Acad Sci U S A 1983;41:1735–7.
16. Hayaishi O, Ueno R, Onoe H, et al. Prostaglandin D2
induces sleep when infused into the cerebral ventricle of conscious monkeys. Adv Prostag Thromb Leukot Res 1987;17B:946–8.
17. Hayaishi O. Sleep-wake regulation by prostaglandins D2
. J Biol Chem 1988;263:14593–6.
18. Onoe H, Kim K, Hayaishi O. Prostaglandin D2
as an endogenous sleep-inducing factor in mammalian brain. In: Inoue, S, Krueger JM, eds. Endogenous sleep factors. The Hague: Academic Publishing, 1990:69–76.
19. Urade Y, Hayaishi O. Biochemical, structural, genetic, physiological, and pathophysiological features of lipocalin-type prostaglandin D synthase. Biochim Biophys Acta 2002;1482:259–71.
20. Murphy PJ, Badia P, Myers BL, et al. Nonsteroidal anti-inflammatory drugs affect normal sleep patterns in humans. Physiol Behav 1994;55:1063–6.
21. Naito K, Osama H, Ueno R, et al. Suppression of sleep by prostaglandin synthesis inhibitors in unrestrained rats. Brain Res 1998;453:329–36.
22. Matsumura H, Honda K, Goh Y, et al. Awakening effect of PGE2
in freely moving rats. Brain Res 1989;481:242–9.
23. Matsumura H, Honda K, Choi WS, et al. Evidence that brain prostaglandin E2
is involved in physiological sleep-wake regulation in rats. Proc Natl Acad Sci U S A 1989;86:5666–9.
24. Nishino S, Mignot E, Fruhstorfer B, et al. Prostaglandin E2
and its methyl ester reduce cataplexy in canine narcolepsy. Proc Natl Acad Sci U S A 1989;86:2483–7.
25. Onoe H. Molecular and neuroanatomical mechanisms of sleep-wakefulness regulation by prostaglandins D2
. Nippon Yakurigaku Zasshi 1998;112:343–9.
26. Urade Y, Hayaishi O. Prostaglandin D2
and sleep regulation. Biochim Biophys Acta 1999;1436:606–15.