Introduction

Orexin-A and orexin-B (also called hypocretin-1 and hypocretin-2) are neuropeptides proteolytically produced from a common precursor that are expressed in a specific population of neurons in the lateral hypothalamic area (Sakurai et al. 1998). Orexin-producing neurons (orexin neurons) send axons throughout the brain with especially dense projections to monoaminergic and cholinergic nuclei in the brainstem, which are regions implicated in the regulation of sleep/wakefulness. The finding of orexin deficiency in narcolepsy patients suggests that this hypothalamic neuropeptide plays a crucial role in regulating sleep/wake states (Sakurai 2007). The implication of orexin in narcolepsy suggests the importance of orexin in the normal regulation of vigilance states.

Because activation of orexin receptors by means of orexin agonists or gene therapy is a promising avenue towards therapeutic trials for narcolepsy and potentially other sleep disorders, it is essential to identify any alteration in the sleep/wake architecture as a result of constitutive increase in orexinergic tone. However, the sleep/wake phenotype resulting from constitutive increased orexinergic tone in the brain has not been thoroughly examined. Although we previously reported some of the pharmacological effects of continuous infusion of orexin-A into the lateral ventricle of rats, we only reported the metabolic parameters (Yamanaka et al. 1999). Effects of orexin administration on vigilance parameters have been examined thus far only by bolus administration (Huang et al. 2001; Piper et al. 2000). To examine the effects of chronically increased orexinergic tone in mice, we made transgenic mouse lines that exhibit widespread constitutive expression of a rat prepro-orexin transgene driven by a β-actin/cytomegalovirus hybrid promoter (CAG/orexin transgenic mice; Mieda et al. 2004). We previously reported that this transgene completely prevented cataplectic arrests and other abnormalities of rapid eye movement (REM) sleep of orexin/ataxin-3 narcoleptic mice in which endogenous orexin neurons are absent (Mieda et al. 2004). We also reported that these mice are resistant to fat diet-induced obesity and show increased insulin insensitivity by promoting energy expenditure (Funato et al. 2009). However, sleep abnormality of these mice have not been reported in detail so far. Here, we describe the sleep/wake phenotype of the CAG/orexin transgenic mice in detail. These mice exhibit sleep abnormalities with fragmentation of non-REM sleep and a specific reduction in REM sleep. EEG/EMG studies also reveal incomplete REM sleep atonia with abnormal myoclonic activity during this sleep stage. These results suggest that CAG/orexin mice have an inability to maintain sleep states and reveal the physiological importance of inhibitory controls of the orexin system in the regulation of sleep/wakefulness states and muscle tone especially during REM sleep. The sleep abnormalities in these mice with orexin overexpression suggests that relatively short half-life is desirable for orexinergic agonists developed for the treatment of narcolepsy or other sleep disorders as unwanted abnormalities in sleep or muscle tone may occur with prolonged increased orexinergic tone. Gene therapy approach with constitutive expression of orexin will cause sleep fragmentation. Methods with proper regulation of orexin expression according to vigilance states will be required for such an approach.

Materials and Methods

Animal Care

All experimental procedures involving animals were approved by the University of Tsukuba Animal Care and Use Committee, the Kanazawa University Animal Care and Use Committee, and the Institutional Animal Care and Research Advisory Committee of the University of Texas Southwestern Medical Center at Dallas and were in accordance with the NIH guidelines.

CAG/orexin Transgenic Mice

To distinguish endogenous mouse orexin gene and transgene, we used rat prepro-orexin cDNA (585 bp) as a transgene, which was inserted between two EcoRI sites of pCAGGS (Niwa et al. 1991). The transgene was excised by digesting with SalI and PstI, gel-purified, and then injected into pronuclei of fertilized eggs obtained from DBF1 mice. We obtained six transgenic lines. Copy numbers of the transgene in these lines were 1, 3, 3, 3, 12, and 12 as determined by genomic Southern hybridization analysis (data not shown). In the present study, we used the lines L2 and L62 for analyses, which have copy numbers of 3 and 12, respectively (Table 1). All experiments utilized male and female mice from crosses in which hemizygous transgenic mice were bred to wild-type C57BL/6J mice. Offspring (DBA1 backcrossed to C57BL/6J, N5–N7) were weaned at 21 days and genotyped by PCR using the following primers: 5′-GTTGTGCTGTCTCATCATTTTGGC-3′ and 5′-AACGTCTTTATTGCCTAGGGACCG-3′.

Table 1 Expression levels of orexin mRNA and peptide in six CAG/orexin transgenic lines
Full size table

Expression of Orexin in the CAG/orexin Mice

Expression levels of orexin mRNA and peptide levels in the brains of transgenic lines were examined by Northern blot hybridization analysis and radioimmunoassay, respectively. For Northern blot hybridization, total RNA from the brains of 4-week-old transgenic mice were isolated, separated by electrophoresis on a 1.0% agarose/formaldehyde gel, blotted onto nylon membranes, and hybridized with a rat α[32P]-labeled prepro-orexin cDNA probe. Membranes were subjected to autoradiography and rehybridized with a glyceraldehyde-3-phosphate dehydrogenase (GAPDH) probe. Relative content of prepro-orexin mRNA was determined by densitometry and calculated as a ratio of orexin/GAPDH. For radioimmunoassay, brains from mice 4 weeks of age for each line were homogenized with a Polytron in 70% (v/v) acetone, 1 M acetic acid, and 20 mM HCl. The homogenate was centrifuged at 20,000×g for 30 min. The resultant supernatant was collected and extracted twice with diethyl ether. The supernatant was loaded onto SepPak C18 cartridges (Waters), which were pre-equilibrated with 0.1% (v/v) trifluoroacetic acid (TFA). Cartridges were washed with 5% CH3CN/0.1% TFA and then eluted with 50% CH3CN/0.1% TFA. The eluate was lyophilized and subjected to radioimmunoassay using commercially available [125I]RIA kits (Phoenix Pharmaceuticals).

Sleep Studies

EEG/EMG monitoring and analysis was performed as previously described (Chemelli et al. 1999; Hara et al. 2001). For quantifying abnormal motor activity during REM sleep, all REM sleep events during a 24-h period were scored (three animals per genotype). Phasic muscle activity during REM sleep was defined as an EMG deflection >50% above the baseline. A tonic activity event during REM sleep was scored when a series of deflections lasted for longer than 0.3 s. Statistical comparisons were performed by ANOVA by post hoc analysis of significance by Fisher’s protected least significant difference test.

Histological Analysis

Mice were anesthetized by intraperitoneal injection of sodium pentobarbital and perfused via the heart with phosphate-buffered saline (PBS), followed by 0.1 M phosphate buffer containing 4% paraformaldehyde. The whole brain was post-fixed in 4% paraformaldehyde in PBS overnight at 4°C. Cryostat sections (40-μm thick) were preincubated with a solution of 1% bovine serum albumin diluted in 0.3% Triton X-100 in PBS for 1 h and then incubated overnight at 4°C with goat anti-ChAT antibody (Chemicon; 1:200) and rabbit anti-orexin antiserum (1:1,000; Nambu et al. 1999) diluted in a fresh hybridization solution. After washing three times in 0.3% Triton X-100 in PBS, the sections were incubated with Alexa Fluor 488-conjugated donkey anti-rabbit IgG (Molecular Probes; 1:250) for 90 min at room temperature and then washed three times in 0.3% Triton X-100 in PBS. Tissues were subsequently incubated with either an Alexa Fluor 594-conjugated donkey anti-goat IgG or goat anti-mouse IgG (Molecular Probes; 1:250) (Cappel) for 90 min at room temperature. The sections were then washed three times in 0.3% Triton X-100 in PBS and mounted onto glass slides and coverslipped. Slides were examined under a fluorescence microscope. Specificity of all antibodies was confirmed using a preabsorption with each respective substance and observing no signal.

Results

Independent Lines of CAG/orexin Transgenic Mice

Immunohistochemical analysis of the brains of one of the transgenic lines, namely, L2, was previously reported (Mieda et al. 2004). Patterns of orexin expression in the brains of each line were similar among each line (data not shown). The absolute amounts of brain orexin expression (as determined by Northern blots) and peptide levels of orexin-A and orexin-B (as determined by orexin-A RIAs) varied with each transgenic line (Table 1).

In this study, we utilized two lines, L2 and L62, which show moderate and high expression levels, respectively (Table 1). Because a subtle gene–dosage effect was observed in parameters of vigilance states between the L2 and L62 lines, the data are reported separately for each line.

Sleep/Wakefulness Abnormalities in CAG/orexin Mice

To examine the effects of orexin overexpression on spontaneous sleep/wake patterns, we monitored the EEG/EMG of mice from two independent lines (L2 and L62) of CAG/orexin transgenic mice and their wild-type littermates. Hypnograms from individual mice of each genotype during the light and dark phases illustrate the behavior of these animals (Fig. 1). Compared with wild-type controls, CAG/orexin mice exhibited evidence of disturbed sleep especially during the light phase (rest period). In particular, CAG/orexin mice have increased fragmentation of sleep patterns during the light phase when normal mouse behavior is predominantly sleep. Increases in the number of wakefulness episodes and decreases in the duration of non-REM and REM sleep support this observation (Fig. 2). Most notably, normal patterns of REM sleep are profoundly disturbed in transgenic mice during both phases. CAG/orexin mice showed significant reduction in the amount of REM sleep during the light phase and significantly reduced duration of REM sleep time across light and dark phases (Fig. 2). Additionally, CAG/orexin mice exhibited reduced REM latencies: REM latencies of wild-type mice, CAG L2 mice, and CAG L62 mice were 9.05 + 0.64 min, 6.17 + 0.61 min (p < 0.05), and 6.19 + 0.17 min (p < 0.05), respectively, for the light period and 8.33 + 0.63 min, 5.42 + 0.92 min (p < 0.05), and 4.92 + 0.53 min (p < 0.05), respectively, for the dark period. These observations strongly suggest that chronically increased orexinergic tone disrupts normal sleep architecture.

Fig. 1
figure 1

Illustrative hypnograms from wild-type and CAG/orexin transgenic mice. Sleep/wake patterns (W wakefulness, N non-REM sleep, R REM sleep) generated from concatenating EEG/EMG stage scores over the first 6 h of the light phase are shown. Note the increased fragmentation of non-REM sleep and reduced amounts and durations of REM sleep in transgenic mouse. The hypnograms of the CAG/orexin mouse shown is from line L2

Full size image
Fig. 2
figure 2

Sleep/wakefulness parameters recorded from two CAG/orexin transgenic lines and wild-type controls. Total time spent in each state (in minutes; a), episode duration (in seconds; b), and episode frequency (per hour; c) over light and dark periods (left and right panels, respectively) are shown. Statistical analysis was tested by repeated-measures ANOVA followed by Tukey post hoc test. *p < 0.05, significant differences of transgenic groups compared with the wild-type group

Full size image

Abnormalities of Muscle Tonus in CAG/orexin Mice during REM Sleep

Close observation of the structural features of EEG/EMG patterns reveals other abnormalities in CAG/orexin mice as well. The onset of REM sleep in normal wild-type mice is characterized by REM sleep atonia, a further reduction of postural muscle tone compared with that preceding non-REM sleep. Although twitching of peripheral muscles (e.g., hands, feet, facial musculature) often occurs during REM sleep, postural muscles remain virtually quiet. In contrast, CAG/orexin mice exhibited sporadic increases in postural muscle tone with the onset of REM sleep. A striking increase in the frequency and amplitude of phasic twitches recorded from nuchal muscles was also observed (p = 0.05; Fig. 3a, b).

Fig. 3
figure 3

Abnormalities of REM sleep atonia in CAG/orexin transgenic mice. a Illustrative EEG/EMG traces reflect REM sleep in wild-type and transgenic mice. Note increased postural tone and increased phasic activity (arrows) evident in the EMG of the transgenic mouse. b Frequencies of tonic and phasic activities in wild-type, CAG/orexin, and orexin/ataxin-3 transgenic mice during REM sleep. Note high frequencies of abnormal tone and phasic activity in CAG/orexin mice as well as a milder elevation of phasic activity in orexin/ataxin-3 transgenic mice

Full size image

Interestingly, abnormalities of atonia in REM sleep have reportedly been associated with narcolepsy in humans. Therefore, for comparison, we also examined narcoleptic orexin/ataxin-3 transgenic mice (Hara et al. 2001) on a similar genetic background as the CAG/orexin mice. Notably, orexin/ataxin-3 mice also exhibited a mild increase in phasic muscle activity during REM sleep, although it was of an intermediate frequency relative to wild-type and CAG/orexin transgenic mice (Fig. 3b). On the other hand, orexin/ataxin-3 mice showed normal frequencies of tonic muscle activity during REM sleep.

These observations suggest that instability of the muscle tone regulatory system in REM sleep is disrupted in both orexin-deficient and overexpressing mice, suggesting that orexins both positively and negatively affect muscle tone during REM sleep.

Discussion

Our results demonstrate that constitutive and ectopic increase of orexinergic transmission results in complex alterations of sleep/wakefulness states. Primarily, orexin overexpression induced an inability to maintain sleep states as evidenced by increased number of wakefulness and non-REM sleep episodes across light and dark phases and a selective decrease in the duration of non-REM sleep episodes during the light phase. These findings contrast with similar studies conducted on mice with orexin-deficient narcolepsy in which animals are primarily unable to maintain wakefulness states during the dark phase (Chemelli et al. 1999; Hara et al. 2001). Strong disruption of normal sleep states was also evident in the CAG/orexin transgenic mice (Figs. 1 and 2). While narcoleptic orexin-deficient mice display frequent and abnormal intrusions of REM sleep-related phenomena into wakefulness, CAG/orexin transgenic mice displayed a significant decrease in the amount of REM sleep during the light phase and reduced REM sleep duration across both light and dark phases. These findings are consistent with a role of orexin in the inhibition of REM sleep. However, reduced REM latency and REM intervals in the CAG/orexin transgenic mice also suggests that maintenance of non-REM sleep is also disturbed by intrusion of REM sleep as well as wakefulness.

The finding that orexin imbalance also disrupted mechanisms of REM sleep atonia may have particular clinical relevance for disorders of abnormal muscle activity and myoclonus during sleep such as restless legs syndrome (RLS) or REM sleep behavior disorder (Schenck and Mahowald 2002). Indeed, one study indicated that early-onset familial forms of RLS are associated with higher levels of orexin-A in the cerebrospinal fluid (Allen et al. 2002). In this study, we identified increases in phasic muscular activity during REM sleep in both orexin overexpressing mice as well as in orexin neuron-ablated orexin/ataxin-3 mice. A potentially complex role of orexins in such processes is also indicated by orexin microinjection studies in cats that suggest that REM sleep-related muscle atonia can be facilitated or inhibited depending on the site of injection in brainstem areas that have been previously implicated in the control of REM sleep atonia (Mileykovskiy et al. 2002). Muscle tone during REM sleep is inhibited via an activation of cholinergic neurons in the pedunculopontine/laterodorsal tegmental nucleus (PPN/LDT). These cholinergic neurons activate glycinergic cells that send inhibitory projections to motor neurons of postural muscles. Orexin neurons send abundant projections to the PPN/LDT (Supplementary Figure 1). Therefore, orexins might regulate the activity of these cholinergic cells, and this mechanism might be important in the pathophysiology of cataplexy in narcolepsy. A study suggests that orexins inhibit cholinergic neurons in the PPN via an activation of local GABAergic interneurons, although orexins have also been shown to evoke direct excitatory effect on the cholinergic neurons in cats (Takakusaki et al. 2005). Constitutive elevation of orexinergic tone in the brain of the CAG/orexin mice might disturb the normal muscle tone regulatory mechanism in LDT/PPN regions during REM sleep. Interestingly, orexin/ataxin-3 mice, in which orexin neurons are genetically ablated, also show a mild increase in phasic muscle activity during REM sleep. These observations that either an increase or decrease in orexinergic tone increases phasic muscle activity during REM sleep suggest an intriguing possibility. Although recent studies show that orexin neurons are active during wakefulness and silent during both non-REM and REM sleep (Lee et al. 2005; Mileykovskiy et al. 2005), some populations of orexin neurons might send direct excitatory projections to cholinergic cells and could be active during REM sleep to keep muscle tone silent. On the other hand, a different population of orexin neurons might send inhibitory signals to these cholinergic cells via an activation of GABA interneurons and/or monoaminergic neurons.

The sleep/wake phenotype of the CAG/orexin mice is totally different from that of mice or rats with pharmacological administration of orexins, which is characterized by prolonged wakefulness. The difference might possibly stem from chronic alterations by plasticity of neurons which express orexin receptors.

Our present findings suggest that the activity of orexin neurons should be appropriately decreased or turned off during sleep to maintain normal sleep. This regulation might possibly be through GABAergic inhibition by the innervation from the sleep-active neurons in the preoptic areas (Sakurai et al. 2005; Yoshida et al. 2006). Our present results also suggest that, if orexinergic agonists are developed as a therapeutic agent for narcolepsy, a relatively short half-life is required in order to avoid any possible unwanted sleep or muscle tone disturbances. Finally, careful regulation of orexinergic tone could potentially be of therapeutic benefit not only for narcoleptic patients but also for patients suffering from abnormal muscle activity and myoclonus during sleep.