Narcolepsy susceptibility gene CCR3 modulates sleep-wake patterns in mice

Narcolepsy is caused by the loss of hypocretin (Hcrt) neurons and is associated with multiple genetic and environmental factors. Although abnormalities in immunity are suggested to be involved in the etiology of narcolepsy, no decisive mechanism has been established. We previously reported chemokine (C-C motif) receptor 3 (CCR3) as a novel susceptibility gene for narcolepsy. To understand the role of CCR3 in the development of narcolepsy, we investigated sleep-wake patterns of Ccr3 knockout (KO) mice. Ccr3 KO mice exhibited fragmented sleep patterns in the light phase, whereas the overall sleep structure in the dark phase did not differ between Ccr3 KO mice and wild-type (WT) littermates. Intraperitoneal injection of lipopolysaccharide (LPS) promoted wakefulness and suppressed both REM and NREM sleep in the light phase in both Ccr3 KO and WT mice. Conversely, LPS suppressed wakefulness and promoted NREM sleep in the dark phase in both genotypes. After LPS administration, the proportion of time spent in wakefulness was higher, and the proportion of time spent in NREM sleep was lower in Ccr3 KO compared to WT mice only in the light phase. LPS-induced changes in sleep patterns were larger in Ccr3 KO compared to WT mice. Furthermore, we quantified the number of Hcrt neurons and found that Ccr3 KO mice had fewer Hcrt neurons in the lateral hypothalamus compared to WT mice. We found abnormalities in sleep patterns in the resting phase and in the number of Hcrt neurons in Ccr3 KO mice. These observations suggest a role for CCR3 in sleep-wake regulation in narcolepsy patients.


Introduction
Narcolepsy is a sleep disorder characterized by excessive daytime sleepiness, fragmented nocturnal sleep, and abnormal rapid eye movement (REM) sleep symptoms (cataplexy, hypnagogic hallucination, and sleep paralysis). Although healthy individuals start nocturnal sleep with a non-REM (NREM) episode, narcolepsy patients show sleep-onset REM periods [1].

Ethics statement
The experimental procedures were approved by the ethics committee on animal experiments at Tokyo Metropolitan Institute of Medical Science and conducted according to the "Guidelines for Care and Use of Laboratory Animals" of the National Institutes of Health.

Surgery
Eight-week-old male mice were anesthetized with pentobarbital (50 mg/kg, i.p.), and implanted with cortical electroencephalogram (EEG) electrodes and nuchal electromyography (EMG) electrodes, which was modified from previous study [30]. In brief, two silver balled wires (diameter, 0.8 mm) were symmetrically placed over the cortex (1.5 mm posterior to bregma, 2.0 mm from midline) for EEG recordings, and two stainless steel wires (diameter, 0.1 mm) were embedded in the neck muscle for EMG recordings. After surgery, mice were individually habituated to the recording environment for at least 10 days in chambers at a constant temperature (25 ± 1˚C) and a 12:12-h light-dark cycle with lights on from 8:00 A.M. to 8:00 P. M., corresponding to Zeitgeber time (ZT) 0-12. All mice received antibiotics immediately after surgery. All mice were monitored for signs of discomfort daily for 10 days after surgery and administered analgesic if necessary. Mice that lost 20% or more of their body weight or had markedly reduced mobility were humanely sacrificed.

Immunohistochemistry
Eleven-week-old male mice (WT, n = 5; Ccr3 KO, n = 5) were deeply anesthetized with pentobarbital in the middle of the light phase (ZT6, mice were motionless with their eyes closed) and perfused with 20 ml of cold PBS followed by 15 ml of 4% paraformaldehyde (PFA) in PBS. The brains were dissected coronally from bregma to 3.5 mm posterior to bregma using a brain slicer. Isolated brains were post-fixed in 4% PFA in PBS at 4˚C for 6 h, cryo-protected in 20% sucrose in PBS at 4˚C for at least 2 days, and frozen in OCT compound (Sakura Finetek, Tokyo, Japan). Floating sections were cut at 40 μm using a cryostat (Leica Biosystems, Wetzlar, Germany) and stored in 0.02% sodium azide in PBS at 4˚C. Sections were incubated for 30 min in 0.1% Triton X-100 in PBS (PBST) containing 0.03%H 2 O 2 to inactivate endogenous peroxidases and then incubated in blocking buffer (5% bovine serum albumin in PBST) at room temperature (RT) for 1 h. After blocking, sections were incubated in rabbit anti-c-

Statistical methods
The primary endpoint of this study was to determine the effect of genotype (WT, Ccr3 KO) on each phenotype. For the time spent in each behavioral state, sleep structure, the number of states transition, and the number of cells, the two-tailed Student's t test was used to compare between means of two unpaired groups. A P-value < 0.05 was considered statistically significant.

Sleep-wake behavior in Ccr3 KO mice
To investigate the role of CCR3 in the development of narcolepsy, we examined sleep-wake patterns in Ccr3 KO mice. Both Ccr3 KO and WT mice exhibited normal nocturnal patterns. Although the percentages of time spent in wakefulness, REM sleep, and NREM sleep in the dark phase were not significantly different between the two genotypes, these sleep-wake patterns in Ccr3 KO mice in the light phase were significantly different compared to WT mice: the proportion of time spent in wakefulness was higher in Ccr3 KO mice (42.9 ± 6.8% in Ccr3 KO vs. 36.1 ± 3.1% in WT, P = 0.019), and the proportion of time spent in REM sleep (3.1 ± 1.2% in Ccr3 KO vs. 4.8 ± 1.3% in WT, P = 0.013) and NREM sleep (54.0 ± 6.4% in Ccr3 KO vs. 59.1 ± 2.3% in WT, P = 0.047) was lower in Ccr3 KO mice compared to WT mice (Fig 2). Sleep continuity in the light phase was affected in Ccr3 KO mice. The number of wakefulness bouts was significantly greater in Ccr3 KO mice (184.9 ± 37.6 times in Ccr3 KO vs. 139.1 ± 22.7 times in WT, P = 0.009), whereas the duration of each wakefulness bout did not differ from WT mice (Fig 3). The number of REM sleep bouts was fewer in Ccr3 KO mice (20.7 ± 6.6 times in Ccr3 KO vs. 32.0 ± 14.1 times in WT, P = 0.047). Interestingly, the number of NREM sleep bouts was significantly greater in Ccr3 KO mice (186.6 ± 38.8 times in Ccr3 KO vs. 143.8 ± 25.0 times in WT, P = 0.018), and the duration of each NREM bout in Ccr3 KO mice was significantly shorter than that in WT mice (139.3 ± 37.6 sec in Ccr3 KO vs. 203.4 ± 47.6 sec in WT, P = 0.007) (Fig 3). These findings indicate that the increased number of NREM bouts and the shorter NREM sleep duration in Ccr3 KO mice are due to the increased number of wakefulness bouts, which suggests fragmented sleep. The calculated number of transitions between different behavioral states in the light phase was greater in Ccr3 KO mice (391.4 ± 74.2 times in Ccr3 KO vs. 313.9 ± 49.9 times in WT, P = 0.025) (Fig 4A). The number of transition between wakefulness and NREM sleep in the light phase was significantly increased in Ccr3 KO mice compared to WT mice (wakefulness to NREM sleep, 183.8 ± 37.7 times in Ccr3 KO vs. 138.4 ± 22.9 times in WT, P = 0.001; NREM sleep to wakefulness, 165.7 ± 41.6 times in Ccr3 KO vs. 111.5 ± 28.3 times in WT, P = 0.007) (Fig 4B). The number of transition among other states was not significantly different between the two genotypes.
In contrast to the light phase, LPS administration significantly decreased the proportion of time spent in wakefulness and increased the proportion spent in NREM sleep in the dark Altered sleep-wake patterns in CCR3 KO mice phase in both Ccr3 KO and WT mice, but similar to before LPS administration, the proportion of time spent in each behavioral state in the dark phase after LPS administration was not different between the two genotypes (Fig 7).

Decreased Hcrt neurons in Ccr3 KO mice
Since Hcrt deficient mice reportedly exhibit fragmented sleep patterns [6,7,33], we quantified the number of Hcrt neurons in Ccr3 KO mice. Ccr3 KO mice had significantly fewer HCRT immunoreactive cells compared to WT mice (3497.4 ± 135.2 in Ccr3 KO vs. 3963.8 ± 149.4 in WT, P = 0.0009) (Fig 8). No significant genotype-dependent differences were observed in the number of FOS-immunoreactive Hcrt neurons (S1A and S1B Fig). We also quantified prepro-Hcrt mRNA levels and hypocretin-1 peptide levels in the hypothalamus (S1 Methods). Neither the mRNA nor the peptide levels were significantly different between the two genotypes (S2 and S3 Figs).

Discussion
In the present study, we showed that Ccr3 KO mice exhibit fragmented sleep patterns in the light phase (Figs 2-4). To our knowledge, this is the first report showing that CCR3 is involved in the sleep regulation. We also observed a slight but significant decrease in the number of Hcrt neurons (Fig 8).
The sleep architecture of Hcrt deficient mice has been thoroughly investigated using prepro-Hcrt KO mice [6] and Hcrt neuron-ablated (Hcrt/ataxin3) mice [7]. These mice exhibit Altered sleep-wake patterns in CCR3 KO mice cataplexy, and direct transitions to REM sleep from wakefulness were observed in both studies. Fragmentation of wakefulness in the dark phase was also demonstrated: a shared phenotype of Altered sleep-wake patterns in CCR3 KO mice a shorter duration of wakefulness episodes was observed in Hcrt deficient mice compared to WT mice [6,7]. A recent study using conditional Hcrt neuron-ablated mice, in which Hcrt neurons can be manipulated to degenerate during adulthood, reported that fragmented sleepwake cycles occur after loss of 80% of Hcrt neurons, and cataplexy is detected after loss of 95% of Hcrt neurons [33]. They also showed that the transition frequency between sleep-wake states is significantly increased in Hcrt deficient mice only in the dark phase, and not in the light phase. In our study, the number of transitions between behavioral states was increased in Ccr3 KO mice only in the light phase, although a similar tendency was observed in the dark phase (Fig 4A). Concurrent with this observation, the differences in sleep structure between Ccr3 KO and WT mice were observed only in the light phase, except for the shorter duration of NREM sleep bouts in Ccr3 KO mice in the dark phase (Figs 5 and 7). The phenotypes of Ccr3 KO mice were different from those of Hcrt deficient mice; however, the fragmented Our previous study reported that impaired CCR3 function is a risk for narcolepsy [17]; yet, a defect only in CCR3 is not sufficient for the disease development because polygenic risks for narcolepsy have been estimated to explain 58.1% of narcolepsy onset in the context of the Values are mean ± SE of 8 WT and 9 Ccr3 KO mice. The two-tailed unpaired Student's t-test was used to compare means between two unpaired groups. *P < 0.05; **P < 0.01. https://doi.org/10.1371/journal.pone.0187888.g007 Altered sleep-wake patterns in CCR3 KO mice contribution of common variants [34]. Therefore, we sought to address whether some triggers besides genetic factors can affect sleep-wake patterns in Ccr3 KO mice. Administration of LPS increased wakefulness and decreased NREM sleep in both Ccr3 KO and WT mice in the light phase (Fig 5), but the changes induced by LPS administration were significantly larger in Ccr3 KO mice compared to WT mice in the light phase (Fig 6). We also found a tendency in which LPS administration induced more fragmented sleep patterns in Ccr3 KO mice in the light phase compared to WT mice (Fig 5). In contrast to the light phase, no significant genotypedependent differences were observed in the phenotypes during the dark phase regardless of LPS administration. The sleep-wake state is regulated by a cluster of neurons in the hypothalamus. Sleep-promoting neurons located in the ventrolateral preoptic nucleus (VLPO) are dominant in the resting phase, whereas wake-promoting neurons in the tuberomammillary nucleus and in the lateral hypothalamus are active during vigilance [35]. Our results suggest that CCR3 promotes NREM sleep in the sleep-promoting system that is dominant in the resting phase. Accordingly, the resting-phase sleep state of Ccr3 KO mice may be easily unstabilized by external stimuli (LPS in this study).
The effects of LPS administration on sleep have been examined in several studies [25][26][27][28][29]. In the dark phase, a shared phenotype of an increased NREM sleep is observed regardless of experimental condition. LPS also reportedly induces fragmented sleep [25,29]: the increased Altered sleep-wake patterns in CCR3 KO mice number of NREM bouts and the shorter NREM sleep duration. Our study replicated these previous findings (Fig 7), suggesting the validity of our results. The result of no genotype-dependent differences in the phenotypes indicates that CCR3 plays a minor role in the arousal state.
The LPS-induced phenotypes observed in the light phase are reportedly different depending on the timing of administration [28]: administration at dark onset partially suppresses REM sleep with no effects on NREM sleep, but administration at light onset completely suppresses REM sleep and promotes NREM sleep. In our study, LPS administration decreased NREM sleep in the light phase. One possible explanation is that LPS administration induces inflammatory hyperalgesia [36] that might result in acute suppression of NREM sleep. We suppose that Ccr3 KO mice are more vulnerable to such disruption in regulation of NREM sleep in the light phase compared to WT mice.
Based on these findings, we hypothesize that CCR3 promotes NREM sleep in the resting phase, and thus CCR3 deficiency is more likely to cause NREM sleep dysregulation after exposure to environmental stimuli that results in fragmented sleep. More studies are needed to understand how CCR3 contributes to the regulation of NREM sleep. For instance, it is crucial to examine whether Ccr3 is expressed and functions in the VLPO. Limitation of our study is that we used only LPS. An investigation using other immune stimulators such as viral single/ double-stranded DNA/RNA is essential for understanding the interactive contribution of CCR3 and environmental factors to the narcolepsy development.
We found that Ccr3 KO mice have~10% fewer Hcrt neurons compared to WT mice. This small decrease alone is not expected to directly induce fragmented sleep because loss of 80% of Hcrt neurons is necessary to cause fragmented sleep-wake cycles that result in narcolepsy [33]. Furthermore, Ccr3 KO mice did not show cataplexy-like behaviors or sleep-onset REM periods, which are typical phenotypes of narcolepsy. According to Tabuchi et al., cataplexy-like behaviors in Hcrt deficient mice are strongly dependent on the Hcrt system. Given these previous reports and our results, fragmented sleep-wake patterns in Ccr3 KO mice in the light phase may be mediated by alternative unknown mechanisms other than the Hcrt system. However, CCR3 may be necessary for survival of Hcrt neurons. Links between neuronal cell survival and CCR3 have been described: Ccr3-expressing CD4+ T cells are necessary for facial motoneuron survival after facial nerve axotomy [37], whereas CCR3 promotes neuronal cell death in a post-infarction mouse model [38]. These reports suggest that the role of CCR3 in neuronal survival is different depending on the context of disease. As genetic and epidemiologic studies have suggested, a loss of Hcrt neurons in patients with narcolepsy is considered to be caused by an autoimmune-mediated process [39]. Multiple studies using sera and cerebrospinal fluids have revealed abnormalities in the levels of cytokines in patients with narcolepsy [40][41][42][43]. Despite evidence indicating abnormalities in immunity, little is known about the state of the immune system during the process of degeneration of Hcrt neurons in narcolepsy patients. A recent report suggested that Th1-related cytokines tend to be increased in narcolepsy patients very close to disease onset [44]. The cytokines that are elevated include chemokine (C-C motif) ligand 5 (CCL5), a ligand for CCR3. Another study showed that intramuscular injection of AS03 alone induces transient expression of chemokines including CCL5 in the injection site of mice [45]. Given that AS03 adjuvanted vaccinations increase the risk for narcolepsy, we speculate that elevated chemokines may mediate the immune response that results in the increased risk of narcolepsy; meanwhile, the CCL5 receptor, CCR3, plays a protective role in Hcrt neuronal survival.
In conclusion, we propose that impaired function of CCR3 causes fragmented sleep that can contribute to vulnerability to developing narcolepsy in an inflammatory context induced by environmental triggers. Understanding the context of narcolepsy onset using animal models of immune-mediated degeneration of Hcrt neurons will be crucial.