Involvement of granulocyte–macrophage colony-stimulating factor (GM-CSF) in pregnancy-enhanced sleep
Abstract
Abstract Granulocyte–macrophage colony-stimulating factor (GM-CSF) is a crucial cytokine for establishing pregnancy. It has been demonstrated previously in rats that sleep increases during early pregnancy and that centrally administered GM-CSF promotes both rapid eye movement (REM) and non-REM sleep. Therefore, whether GM-CSF is involved in pregnancy-enhanced sleep was investigated using the anti-GM-CSF antibody. Female rats received an intracerebroventricular infusion of either anti-GM-CSF or control IgG (10 μg each) for four nights from the first day of pregnancy (PD1–PD4). Although sleep amounts on PD1 were not affected, anti-GM-CSF decreased non-REM and REM sleep significantly during PD2–PD4 compared with the control baseline of the IgG group. The results demonstrated that anti-GM-CSF treatment suppresses pregnancy-enhanced sleep, suggesting that GM-CSF contributes to sleep regulation during pregnancy.
INTRODUCTION
Granulocyte–macrophage colony-stimulating factor (GM-CSF) was originally described as a haemopoi-etic cytokine, but it also exerts a potent effect on reproduction. The synthesis of GM-CSF is induced after mating, and significant amounts of GM-CSF are released from epithelium cells in the uterus and accessory organs at the beginning of pregnancy.1 Synchronized with the elevated levels of GM-CSF in the periphery during the period of early pregnancy, rapid eye movement (REM) sleep and non-REM sleep are enhanced in rats.2,3 It is also known that GM-CSF crosses the blood–brain barrier.4 Furthermore, we demonstrated recently that GM-CSF promotes both REM and non-REM sleep in the rat when GM-CSF is administered intracerebroventricularly (i.c.v.).5 It is possible that GM-CSF could be responsible for pregnancy-enhanced sleep. To test this hypothesis, we examined whether blocking the central effects of GM-CSF with its antibody suppresses sleep in pregnant rats.
MATERIALS AND METHOD
Female Sprague–Dawley rats, weighing 270–340 g, were used. The rats were caged in groups on a 12-h light/dark cycle (lights on at 08.00 hours). Under pentobarbital sodium (40 mg/kg, intraperitoneally) anaesthesia, the rats were implanted with electroencephalograpy (EEG) and electromyography (EMG) electrodes, a brain thermistor, and an i.c.v. cannula into the third ventricle. After a week of recovery from surgery, the rats were moved to an experimen-tal chamber, housed individually, and then allowed to adjust to the continuous i.c.v. infusion of saline (10 μL/h). Meanwhile, vaginal smears were inspected daily to check their estrous cycle. During a pro-estrous night, the rats were exposed to mating in pairs with male rats. The next day, after confirming the presence of seminal traces in the smears, was designated as pregnant day 1 (PD1). The pregnant rats received an i.c.v. infusion of 10 μg of either anti-mouse GM-CSF antibody (anti-GM-CSF) or normal goat IgG (control IgG) for 10 h during the dark period on PD1–PD4. Polygraphic recordings of EEG, EMG, and Tbr were made for the entire four days from the beginning of the light period on PD1. Differences in the effects of anti-GM-CSF and control IgG on pregnancy-enhanced sleep were compared statistically using two-way ANOVA.
RESULTS
An i.c.v. infusion of control IgG did not alter the basic sleep patterns enhanced by the beginning of pregnancy, although increases in non-REM sleep were affected to a lesser degree and were more prominent in REM sleep than those in a previous study (Table 1).2 An i.c.v. infusion of anti-GM-CSF exerted no significant effect on sleep on PD1. However, anti-GM-CSF treatment decreased the nocturnal amounts of non-REM and REM sleep significantly during the remainder of the early pregnancy. Compared with the baseline of the control IgG group, non-REM sleep and REM sleep decreased to 80.5% and 65.0%, respectively, in the anti-GM-CSF group during PD2– PD4 (Table 1; P < 0.05). None of the pregnant rats aborted after being treated with the i.c.v. infusion of either IgG or anti-GM-CSF.
| PD 1 | PD 2 | PD 3 | PD 4 | |
|---|---|---|---|---|
| Non-REM sleep (min) | ||||
| Control IgG (n = 6) | 246.9 ± 9.2 | 250.5 ± 12.1 | 233.7 ± 12.6 | 230.6 ± 17.7 |
| Anti-GM-CSF (n = 5) | 247.4 ± 11.3 | 189.4 ± 4.6* | 189.2 ± 15.4* | 196.6 ± 35.4 |
| REM sleep (min) | ||||
| Control IgG (n = 6) | 75.7 ± 4.3 | 80.2 ± 4.0 | 70.3 ± 3.7 | 62.6 ± 3.1 |
| Anti-GM-CSF (n = 5) | 61.2 ± 6.2 | 47.8 ± 10.6* | 39.6 ± 6.1* | 51.1 ± 14.9 |
- * Significant differences from control IgG; P < 0.05.
- PD, pregnant day; REM, rapid eye movement.
DISCUSSION
In the case of pregnancy, GM-CSF works as a growth factor and plays a role in the development and attachment of blastocysts. A significant amount of GM-CSF is necessary during the pre-implantation period and is released locally in the maternal body. However, it is possible that any excess amount of GM-CSF may get into the blood stream, or even transported into the brain, and could affect behaviour.4 The present study demonstrated that i.c.v.-administered anti-GM-CSF was capable of suppressing pregnancy-enhanced non-REM and REM sleep. Because GM-CSF is reported to be somnogenic,5 it is reasonable to postulate that GM-CSF is one of the sleep-inducing factors in pregnancy.
Other cytokines that promote sleep in the case of infection have been reported. Although they are inflammatory cytokines, some of them are critical during certain periods of pregnancy as well. Therefore, it is likely that other cytokines as well as GM-CSF also contribute to pregnancy-enhanced sleep. Why do pregnant women get sleepier? This could be controlled under a host–defense mechanism because the cytokines are involved in the regulation of pregnancy-altered sleep.
