Skip to main content
NCBI home page
The Journal of Physiology logoLink to The Journal of Physiology
. 2004 Feb 13;556(Pt 3):935–946. doi: 10.1113/jphysiol.2003.056622

Activation of c-fos in GABAergic neurones in the preoptic area during sleep and in response to sleep deprivation

Hui Gong 2, Dennis McGinty 1,2, Ruben Guzman-Marin 2, Keng-Tee Chew 1, Darya Stewart 3, Ronald Szymusiak 1,3
PMCID: PMC1664995  PMID: 14966298

Abstract

Neurones in the median preoptic nucleus (MnPN) and the ventrolateral preoptic area (vlPOA) express immunoreactivity for c-Fos protein following sustained sleep, and display elevated discharge rates during both non-REM and REM sleep compared to waking. We evaluated the hypothesis that MnPN and vlPOA sleep-active neurones are GABAergic by combining staining for c-Fos protein with staining for glutamic acid decarboxylase (GAD). In a group of six rats exhibiting spontaneous total sleep times averaging 82.2 ± 5.1% of the 2 h immediately prior to death, >75% of MnPN neurones that were Fos-immunoreactive (IR) were also GAD-IR. Similar results were obtained in the vlPOA. In a group of 11 rats exhibiting spontaneous sleep times ranging from 20 to 92%, the number of Fos + GAD-IR neurones in MnPN and vlPOA was positively correlated with total sleep time. Compared to control animals, Fos + GAD-IR cell counts in the MnPN were significantly elevated in rats that were sleep deprived for 24 h and permitted 2 h of recovery sleep. These findings demonstrate that a majority of MnPN and vlPOA neurones that express Fos-IR during sustained spontaneous sleep are GABAergic. They also demonstrate that sleep deprivation is associated with increased activation of GABAergic neurones in the MnPN and vlPOA.

The preoptic area (POA) of the hypothalamus is importantly involved in the regulation of sleep–waking states. Electrolytic and neurotoxin lesions of the POA result in persistent insomnia (McGinty & Sterman, 1968; Szymusiak & Satinoff, 1984; Szymusiak et al. 1991; John & Kumar, 1998; Lu et al. 2000). Microinfusion of somnogenic agents into the POA promotes sleep (Ueno et al. 1982; Ticho & Radulovacki, 1991; Mendelson & Martin, 1992). Neurones that exhibit increases in extracellularly recorded discharge rate during sleep compared to waking are localized in the POA (Findlay & Hayward, 1969; Kaitin, 1984; Alam et al. 1995a; Alam et al. 1997).

Neurones in the ventrolateral POA (vlPOA) (Sherin et al. 1996, 1998; Gong et al. 2000) and the median preoptic nucleus (MnPN) (Gong et al. 2000) of rats exhibit immunoreactivity for the c-Fos protein following episodes of sustained sleep but not following sustained waking. Unit recording studies confirm the presence of neurones with elevated discharge rates during both non-REM and REM sleep, compared to waking, in the rat vlPOA (Szymusiak et al. 1998) and MnPN (Suntsova et al. 2002). In the vlPOA, sleep-related c-Fos protein immunoreactivity is colocalized with markers for the inhibitory neuromodulator galanin (Lu et al. 2002; Gaus et al. 2002). Galanin is highly colocalized with gamma-aminobutyric acid (GABA) in vlPOA neurones (Sherin et al. 1996, 1998). The vlPOA is a source of afferents to histaminergic neurones in the tuberomammillary nucleus of the posterior hypothalamus, and to monoaminergic cell groups in the dorsal raphe nucleus and the locus coreleus (Sherin et al. 1998; Steininger et al. 2001). Therefore, sleep-regulatory neurones in the vlPOA can be hypothesized to exert inhibitory modulatory control over multiple monoamine arousal systems during sleep.

The neurotransmitter phenotype of MnPN neurones that exhibit sleep-related Fos-IR is unknown. In the experiments described here, we evaluated the hypothesis that MnPN sleep-active neurones are GABAergic by combining immunostaining for c-Fos protein with immunostaining for glutamic-acid decarboxylase (GAD), a marker of GABAergic neurones. The number of Fos single-, GAD single- and Fos + GAD double-IR neurones was quantified throughout the MnPN and vlPOA in rats exhibiting varying amounts of spontaneous sleep. We also examined patterns of Fos and GAD immunoreactivity in these POA subregions after 24 h of sleep deprivation, to test the hypothesis that increased homeostatic sleep drive is associated with activation of GABAergic neurones in the MnPN and vlPOA.

Methods

All experiments were conducted in accordance with the National Research Council Guide for the Care and Use of Laboratory Animals. All protocols described here were reviewed- and approved by the Animal Care and Use Committee at V.A. GLAHS. Experiments were performed on 35 adult male Sprague–Dawley rats, weighing 280–320 g. Rats were anaesthetized with ketamine/xylazine (80/10 mg kg−1, i.p.) and surgically implanted for chronic cortical electroencephalogram (EEG) and dorsal neck electromyogram (EMG) electrodes for recording sleep, as previously described (Alam et al. 1997). Briefly, stainless steel (ss) screw electrodes were implanted in the skull for EEG recordings and flexible, insulated ss wires were threaded into neck muscles for EMG recording. Leads from these recording electrodes were soldered to a small Amphenol connector and the entire assembly was anchored to the skull with dental acrylic.

All subjects were allowed a 7 day recovery period following surgery. During this time they were housed in a temperature controlled recording chamber (23 ± 0.5°C) and maintained on a 12: 12 light: dark cycle (lights on 08.00 h). In the recording chamber rats had ad libitum access to food and water. For three consecutive days prior to the experimental day, rats were connected to a recording cable that was lightly suspended above them by a counter-weighted beam at 08.00 h. They remained connected to the cable until 11.00 h. The cable joined the miniature connector on the animal's head to a series of AC coupled amplifiers with filters (A-M Systems). The outputs of the amplifiers were directed to a data acquisition board (Data Translation, DT-2821) installed in a Pentium III processor-based PC. The EEG and EMG recordings were digitally displayed and stored continuously to disk using Pass Plus software (Delta Software, St Louis, MO, USA).

Spontaneous sleep

Spontaneous sleep–waking amounts were examined in 11 rats. On the experimental day, rats were connected to the recording cable at 08.00 h. Once connected to the cable, rats were left undisturbed. Continuous EEG and EMG recordings were initiated at 09.00 h and continued for 2 h. At 11.00 h, rats were removed from the recording chamber and immediately given a lethal dose of pentobarbital (100 mg kg−1), which was followed by cardiac perfusion (see below for details) as soon as the animals were deeply anaesthetized.

Sleep deprivation

Twenty-four hour sleep deprivation was achieved with a forced locomotion protocol in a recording chamber/treadmill assembly. This assembly consisted of a plastic chamber (28 cm × 28 cm × 40 cm) that was fixed and suspended 0.3 cm above the vinyl belt of a treadmill, the treadmill belt forming the floor of the cage. Food and water were available at all times while the animals were on the treadmill, and the timing of the light–dark cycle was the same as for the home cage. Sleep deprivation (SD) was achieved by automated activation of the treadmill for 3 s at 15 s intervals (i.e. 3 s on; 12 s off). Activation of the treadmill caused the animal to step to avoid being carried into the wall of the chamber. Activation of the treadmill at 15 s intervals prevented the animal from achieving all but the briefest episodes of light non-REM sleep. The control procedure for sleep deprivation involved subjecting animals to the same amount of treadmill activation over a 24-h period, with the treadmill successively on for 15 min and off for 60 min, allowing the control animals unrestricted sleep for 60 min at 75 min intervals. The effectiveness of SD versus sleep deprived control (SDC) procedures was verified by continuous recordings of EEG and EMG activity while rats were on the treadmill. Between each use, the treadmill belt and cage were scrubbed with detergent, wiped with a bleach solution and rinsed with water.

Rats sbjected to 24 h SD (n= 12) or SDC (n= 12) were adapted to the treadmill apparatus during the 3 days prior to the experimental day, being placed on the treadmill and connected to recording cables from 11.00 h to 15.00 h. The treadmill was activated in the sleep-deprivation mode (3 s on/12 s off) 13.00 h–15.00 h, to allow the animals to acclimate to movement of the treadmill belt. On the experimental day, rats were removed from the home/recording cage at 08.00 h, placed on the treadmill and connected to the recording cables. Continuous recordings of EEG and EMG were initiated and the treadmill was activated in the SD mode or SDC mode.

Following 24 h of sleep deprivation on the treadmill, rats were subjected to one of two postdeprivation conditions. Six SD rats and six SDC rats were permitted 2 h of recovery sleep prior to killing. These rats were removed from the treadmill at 09.00 h, returned to the home recording cage, connected to the recording cable and permitted 2 h of unrestricted sleep from 09.00 h to 11.00 h while EEG and EMG were continuously recorded. At 11.00 h, rats were removed from the chamber, given a lethal dose of pentobarbital followed by perfusion (see below). An additional six rats SD rats and six SDC rats were not permitted recovery sleep prior to killing. The six SD rats were simply left on the treadmill from 09.00 h until 11.00 h the next day (total of 26 h SD). The six SDC rats were also left on the treadmill through CT2, with the treadmill being switched from SDC to SD mode at the start of CT1 (24 h SDC, 2 h SD).

Sleep analysis

For spontaneous sleep conditions, sleep–wake states were scored in 10-s epochs on the basis of the predominant state within each epoch. Sleep–waking state was assessed by a trained scorer from the digitized EEG and EMG recording displayed on the computer monitor by the Pass Plus software. The scorer was blind to experimental condition and to all Fos- and GAD-IR cell count data. Wake was defined by low-voltage, high-frequency activity combined with elevated neck muscle tone. NREM sleep was defined by high-amplitude EEG with prominent activity in the 2- to 4-Hz range. Rapid eye movement (REM) sleep was defined by moderate-amplitude EEG with dominant theta frequency activity (6–8 Hz) combined with minimal neck EMG tonus except for occasional brief twitches. The percentage of each state was calculated for the total 2 h recording period.

Sleep stages were determined during 24 h sleep deprivation and control conditions, and during the 2 h recovery sleep periods, using criteria identical to those used to assess spontaneous sleep. Continuous EEG delta power was computed with the Pass Plus software during recovery sleep (see Fig. 5). The amplified EEG was filtered between 0.5 and 30 Hz and digitized at 256 Hz. Overlapping 4 s Barlett-tapered windows were analysed in 0.5–1.0 Hz bins for EEG frequencies ranging from 0.5 to 30 Hz, and combined in 10 s samples. Spectral power was calculated in the 0.5–4 Hz range for all artifact-free episodes of waking and non-REM and REM sleep. For comparisons of EEG delta power between groups of animals, the ratio of non-REM sleep/waking delta power was calculated as a percentage, and subjected to statistical analysis (Table 2).

Figure 5. Temporal distribution of sleep–waking state progression (lower panels) and continuous EEG delta power (upper panels) during 2 h of recovery sleep in a sleep deprived rat and a sleep deprived control rat.

Figure 5

Open in a new tab

Abbreviations: W, waking; N, non-REM sleep; R, REM sleep.

Table 2.

Percentage time awake and asleep and percentage nonrem/waking EEG delta power during 2 h of recovery sleep and 2 h of spontaneous sleep

24 h Sleep deprived; recovery sleep (n= 6) 24 h Sleep deprived; control recovery sleep (n= 6) High spontaneous sleep (n= 6) F-ratio (2,15)
% Time awake 19.1 ± 5.2 27.1 ± 5.1 17.7 ± 5.1 1.18 NS
% Time non-REM sleep 61.7 ± 6.4 51.9 ± 2.6 69.3 ± 2.9† 4.87 P < 0.05
% Time REM sleep 19.2 ± 3.3 20.9 ± 2.6 13.0 ± 3.4 2.21 NS
% Non-REM/waking EEG delta power 494.5 ± 41.6**†† 358.7 ± 14.9** 272.7 ± 41.5 12.25 P < 0.001
Open in a new tab

†Significantly different from SDC, P < 0.05 (Newman–Keuls). **Significantly different from HSS. P < 0.01 (Newman–Keuls). ††Significantly different from SDC, P < 0.01 (Newman–Keuls).

Immunohistochemistry

Animals were perfused transcardially with 0.1 m phosphate buffered (PB) saline, followed by 300 ml PB 3% paraformaldehyde (pH 7.0) followed by 10% and then 30% sucrose. The brains were then stored in 30% sucrose at 4°C until they sank. Coronal sections were cut at 40 μm on a freezing microtome. For immunohistochemistry, sections were incubated with a rabbit anti-c-Fos primary antiserum (AB-5, Oncogene Science; 1: 20 000) for 48 h. Sections were subsequently incubated with a biotinylated goat antirabbit IgG (Vector Laboratories; 1: 600) for 2 h and then reacted with avidin–biotin complex (ABC, Vector Elite Kit, 1: 200) and developed with Nickel-diaminobenzidine tetrahydrochloride (Ni-DAB), which produced a black reaction product in cell nuclei. Sections were then incubated with rabbit anti-GAD (AB5992 Chemicon; 1: 1000) and developed with DAB to produce a brown reaction product for double labelling. Omission of the GAD primary antiserum resulted in the absence of specific staining.

Cell counts

The Neurolucida computer-aided plotting system (Microbrightfield) was used for identifying and quantifying neurones that were single labelled for c-Fos immunoreactivity, GAD single-labelled neurones and Fos + GAD double-labelled neurones. Counts of each cell type were performed by an individual who was blind to the experimental condition of the animals. Cell counts were performed bilaterally in three consecutive sections containing the largest part of the nucleus under study. Those six counts were then averaged to yield a single value for each rat for each nucleus examined.

Section outlines were drawn under 20× visualization. Fos-immunoreactive (IR), GAD-IR and Fos + GAD-IR neurones were mapped using different symbols and colours in the section outlines under 400× visualization. All cell counts were calculated for constant rectangular grids corresponding to three areas of interest. (1) The rostral MnPN grid was a 600 μm × 600 μm square, centred on the apex of the third ventricle rostral to the decussation of the anterior commissure and to bregma (A: 0.1 mm) (Gong et al. 2000). (2) The caudal MnPN grid was placed immediately dorsal to the third ventricle at the level of the decussation of the anterior commissure, extending 150 μm laterally and 600 μm dorsally just caudal to Bregma (A: –0.26 mm) (Gong et al. 2000). (3) The vlPOA grid was placed such that the ventromedial corner abutted the lateral edge of the optic chiasm and extended 700 μm laterally and 300 μm dorsally into the lateral preoptic area at level 160 μm or more caudal to the organum vasculosum of the lamina terminalis (Scammell et al. 1998).

Statistical analysis

Differences in average total sleep time and cell counts between rats with high spontaneous sleep (HSS) and low spontaneous sleep (LSS) were assessed with independent t tests (Fig. 3). To assess the relationship between spontaneous sleep amounts and activation of GABAergic neurones, a linear regression was calculated for percentage total sleep time and the number of Fos + GAD-IR neurones for all 11 rats studied in the spontaneous sleep protocol (Fig. 4). A one-way non-repeated measures ANOVA was calculated for several variables of interest across the following five conditions: SD, recovery sleep; SDC, recovery sleep; SD, no recovery sleep; SDC, no recovery sleep; and high spontaneous sleep. A P value of < 0.01 was adopted for significance for ANOVAs. Significance of the differences between individual group means was assessed with post hoc tests (Newman–Keuls)

Figure 3. Mean numbers of total Fos-IR neurones, Fos-IR single-labelled neurones, Fos + GAD-IR double-labelled neurones and the mean percentage of Fos-IR neurones that were double labelled for GAD-IR in the rostral MnPN, caudal MnPN and vlPOA for HSS and LSS rats.

Figure 3

Open in a new tab

**P < 0.001, independent t test; *P < 0.01, independent t test.

Figure 4. Regression lines and scatter plots of the number of Fos + GAD-IR double-labelled neurones versus total sleep time in the 11 rats studied in the spontaneous sleep protocol.

Figure 4

Open in a new tab

In each of the three POA subregions examined, there was a significant positive correlation between sleep time and Fos + GAD-IR cell counts.

Results

Spontaneous sleep

The 11 rats studied exhibited spontaneous sleep amounts ranging from 20 to 92% of the 2 h recording period. Spontaneously sleeping rats were divided into two groups based on total sleep time: high spontaneous sleep (HSS; total sleep time ≥60% of the recording period, n= 6) and low spontaneous sleep (LSS; total sleep time <60%, n= 5). Total sleep time averaged 82.2 ± 5.1% (range 61–92%) of the recording period for HSS rats and 34.6 ± 5.8% (range 51–20%) for LSS rats. This difference was statistically significant (t test, P < 0.001).

Figure 1A and B are photomicrographs from the rostral MnPN of one HSS rat and one LSS rat, respectively, showing examples of GAD-IR and Fos + GAD-IR neurones. Fos protein immunoreactivity is stained black and confined to the nucleus. The brown GAD staining is evident throughout the soma and the proximal dendrites. Elimination of the primary GAD antibody during the second incubation in control sections yielded clear nuclear Fos-IR staining, but only diffuse, granular staining in scattered soma. This diffuse staining did not extend into dendrites, and double-labelled profiles as shown in Figure 1A and B were never observed in control sections.

Figure 1. Examples of Fos + GAD-IR double-labelled neurones (filled arrows) and GAD-IR single-labelled cells (open arrows) in the MnPN of one HSS (A) and one LSS (B) rat.

Figure 1

Open in a new tab

Calibration = 20 μm.

Figure 2A contains line drawings through the rostral MnPN from one HSS rat and from one LSS rat. As has been reported previously (Gong et al. 2000), the number of Fos-IR neurones was dramatically elevated in the HSS rat compared to the LSS rat. In addition, a clear majority of the Fos-IR neurones in the HSS rat also stained for GAD (47/58; 81%). While the majority of Fos-IR neurones were GAD-positive in the HSS rat, a significant number of GAD-IR neurones in the rostral MnPN of this rat did not exhibit Fos-IR (35/82; 46%). Similar results were obtained from the same pair of rats for the caudal MnPN (Fig. 2B) and the vlPOA (data not shown).

Figure 2. Number and distribution of Fos-IR single-labelled, GAD-IR single-labelled and Fos + GAD-IR double-labelled neurones located in counting grids for the rostral (A) and caudal (B) MnPN from one HSS rat and one LSS rat.

Figure 2

Open in a new tab

In the HSS rat, 81% of rostral MnPN neurones and 77% of caudal MnPN neurones that stained for Fos-IR were also positive for GAD-IR. •, Fos + GAD double labelled; ▵, GAD single labelled; □, Fos single labelled

Figure 3 compares cell count data for HSS and LSS rats. In all three brain regions studied, the number of total Fos-IR neurones, the number of Fos + GAD double IR neurones, and the percentage of Fos-IR neurones that were double labelled was significantly higher in HSS versus LSS rats. The numbers of Fos single-labelled neurones did not differ between the groups. Figure 3 also demonstrates that in HSS rats, Fos-IR was predominately located in GABAergic neurones, as the percentage of Fos-IR neurones that were double labelled for GAD averaged 80.5 ± 2.5% in the rostral MnPN, 75.7 ± 1.5% in the caudal MnPN and 80.7 ± 0.9% in the vlPOA.

Figure 4 is a scatter plot of percentage total sleep time versus the number of Fos + GAD-IR neurones in the rostral MnPN, caudal MnPN and vlPOA for all 11 rats studied in the spontaneous sleep protocol. In each POA subregion, there was a significant positive correlation between the number of Fos + GAD-IR cells and total sleep time.

Sleep deprivation and recovery sleep

Six animals were subjected to 24 h of SD, followed by 2 h of recovery sleep. An additional six rats were subjected to 24 h SDC procedures, followed by 2 h of recovery sleep at the same times of day (see Methods). Table 1 shows the mean percentage time spent awake, in non-REM sleep and in REM sleep during the 24 h deprivation period for these two groups of rats. Both non-REM and REM sleep amounts were significantly reduced in SD compared to SDC rats (t test, P < 0.001).

Table 1.

Percentage time awake and asleep during 24 h of sleep deprivation and sleep deprivation control procedures

% Waking % Non-REM sleep % REM sleep % Total sleep
24 h Sleep deprived (n= 6) 94.3 ± 2.3 5.5 ± 2.2 0.2 ± 0.1 5.7 ± 2.3
24 h Sleep deprived control (n= 6) 59.6 ± 2.3 36.7 ± 2.1 3.7 ± 0.2 40.4 ± 2.3
Open in a new tab

Shown in Fig. 5 are plots of sleep–waking state progression and of continuous EEG delta power for one SD and one SDC rat during the 2 h recovery sleep period. There is a period of wakefulness at the beginning of the recovery sleep period in both animals, as a result of being handled by the experimenter during transfer from the treadmill to the recording cage. Note, however, that the latency to sleep onset is reduced, that there are fewer awakenings after sleep onset and that EEG delta power during non-REM sleep is elevated in the SD rat compared to the SDC rat.

Table 2 shows mean percentage time spent awake and in non-REM and REM sleep during the 2 h recovery period for the SD and SDC groups, along with the 2 h sleep–wake percentages for the six HSS rats. Mean percentage time awake was somewhat higher in SDC rats compared to the other groups, but ANOVA indicated no significant group effect for this measure. The percentage time in non-REM sleep was significantly lower in SDC versus HSS rats. The difference between SD and HSS rats was not significant. The non-significant increase in percentage time awake and the significant decrease in percentage non-REM sleep time in SDC rats compared to HSS rats reflects the fact that SDC rats were briefly disconnected from recording cables and transferred from the treadmill to a recording chamber at the start of the final 2 h recording, and typically exhibited a > 20 min latency to sleep onset following this transfer (see Fig. 5B). HSS rats were not transferred or otherwise handled immediately prior to the start of the final 2 h recording period. While mean percentage REM sleep time was somewhat elevated in both the SD and SDC conditions, ANOVA indicated no significant group effect for this measure.

There was a highly significant group effect on EEG delta power (Table 2). The mean percentage non-REM/waking delta power was significantly higher in the SD group compared to both SDC and HSS rats. Furthermore, this variable was significantly elevated in SDC rats compared to HSS rats. This latter findings suggests that SDC rats experienced at least a mild degree of sleep deprivation compared to HSS rats.

Fos + GAD-IR cell counts and the percentage of GAD-IR neurones that were also Fos-IR are shown in Figure 6A and B for SD, SDC and HSS rats. In the rostral and caudal MnPN, both the number of Fos + GAD-IR cells and the percentage of GAD+ neurones that were double labelled were significantly higher in the SD group than in the SDC and HSS conditions. In the caudal, but not rostral MnPN, these values were also significantly higher in SDC rats compared to HSS rats. In vlPOA, there were no differences in these variables between SD and SDC rats, but values in both conditions were significantly higher than those in HSS rats.

Figure 6. Mean numbers of Fos + GAD-IR double-labelled neurones (A) and mean percentage of GAD-IR neurones that were double labelled for Fos-IR (B) in SD rats permitted 2 h of recovery sleep (n= 6), in SDC rats permitted 2 h of recovery sleep (n= 6) and in HSS rats (n= 6).

Figure 6

Open in a new tab

ANOVA indicated a significant treatment effect on Fos + GAD-IR cell counts in the rostral MnPN (F2,15= 45.07, P < 0.001), the caudal MnPN (F2,15) = 27.78, P < 0.001) and the vlPOA (F2,15= 15.57, P < 0.001). Similarly, ANOVA indicated a significant treatment effect on the percentage of GAD-IR neurones that were double labelled for Fos-IR in the rostral MnPN (F2,15= 49.65, P < 0.001) caudal MnPN (F2,15= 14.29, P < 0.01) and vlPOA (F2,15= 13.08, P < 0.01). **Significantly different from SD and HSS values, P < 0.01, Newman–Keuls. *Significantly different from HSS values, P < 0.05, Newman–Keuls.

Sleep deprivation without recovery sleep

Six rats were subjected to 24 h of SD, followed by an additional 2 h of SD immediately prior to killing. Six additional rats underwent SDC procedures for 24 h followed by 2 h of SD at the same times of day (see Methods). Table 3 demonstrates that SD and SDC rats experienced the same degree of sleep loss during the final 2 h. The number of Fos + GAD-IR neurones was significantly elevated in SD compared to SDC rats, in the rostral MnPN, caudal MnPN and vlPOA (Fig. 7).

Table 3.

Percentage time awake and asleep during 2 h of sleep deprivation following 24 h SD and SDC

24 h Sleep deprived; no recovery sleep (n= 6) 24 h Sleep deprived control; no recovery sleep (n= 6) F-ratio (2,15)
% Time awake 94.9 ± 2.5 98.7 ± 0.9 2.44 NS
% Time non-REM sleep 2.6 ± 0.8 2.1 ± 0.9 1.51 NS
% Time REM sleep 0.3 ± 0.3 0 1.00 NS
Open in a new tab

Figure 7. Mean numbers of Fos + GAD-IR double-labelled neurones in SD and SDC rats that were not permitted recovery sleep.

Figure 7

Open in a new tab

The cell counts were significantly higher in SD versus SDC rats in the rostral MnPN (F1,10= 17.57, P < 0.01), the caudal MnPN (F1,10= 16.91, P < 0.01) and the vlPOA (F1,10= 15.57, P < 0.01).

Discussion

Neurones expressing c-Fos protein immunoreactivity following sustained sleep are localized in the rat MnPN (Gong et al. 2000) and vlPOA (Sherin et al. 1996), where they are largely segregated from cells that exhibit Fos-immunoreactivity during waking. The MnPN and vlPOA contain high concentrations of neurones that have elevated discharge rates during both non-REM and REM sleep compared to waking, i.e. sleep-active discharge pattern (Szymusiak et al. 1998; Suntsova et al. 2002). We report here that >75% of MnPN and vlPOA neurones that were Fos-IR following sustained spontaneous sleep also stained for GAD. In spontaneously sleeping animals, the number of Fos + GAD-IR neurones in the MnPN and vlPOA was positively correlated with total sleep time. During recovery sleep following sleep deprivation accompanied by increased EEG delta activity, the proportion of MnPN GABAergic neurones that were also Fos-IR was significantly elevated compared to sleep deprivation control animals. Even in the absence of the opportunity for recovery sleep, the number of GABAergic neurones expressing Fos immunoreactivity was significantly elevated in the MnPN and vlPOA of sleep-deprived versus control rats. These findings demonstrate that a majority of putative POA sleep-regulatory neurones are GABAergic and that increased activation of these GABAergic neurones is a potential mechanism by which prior time awake causes increased sleep amount and/or sleep depth.

Neurones in the vlPOA that express sleep-related Fos immunoreactivity were previously shown to immunostain for the inhibitory neuromodulator galanin (Lu et al. 2002; Gaus et al. 2002). Galanin was found to be highly colocalized with GABA in vlPOA neurones (Sherin et al. 1996; Sherin et al. 1998). The present results confirm the implications of these findings with respect to c-fos activation in GABAergic neurones, providing direct evidence of extensive colocalization of sleep-related Fos-immunoreactivity with GAD-immunoreactivity in the vlPOA. The present results provide the first evidence that MnPN neurones expressing Fos-immunoreactivity during sleep are GABAergic. They also demonstrate that putative sleep-regulatory GABAergic neurones are not confined to the vlPOA.

In all of the experimental conditions described here, including spontaneous sleep, sleep deprivation and sleep deprivation control, rats were killed at the same circadian time, i.e. 11.00 h, 3 h after lights-on. The extent to which sleep-related c-fos expression in MnPN neurones is under circadian control remains to be determined.

A possible role for MnPN GABAergic neurones in homeostatic responses to sleep deprivation is supported by the finding of significantly elevated numbers of Fos + GAD-IR neurones following sleep in SD versus SDC and HSS rats. Increased Fos-IR in MnPN GABAergic neurones in SD animals was associated with increased EEG delta power in non-REM sleep, but not with increased total sleep time (Table 2). The absence of significant increases in sleep time in SD rats was due to the fact that animals in all three groups were killed at a time of day when spontaneous sleep propensity was high, i.e. 3 h after lights-on. It has been shown previously that when recovery sleep in rats is recorded during the light portion of the light–dark cycle, prior time awake is correlated with increased non-REM sleep EEG delta power, a widely accepted metric of sleep depth (Tobler & Borbely, 1986), but not with significant increases in sleep amount (Lancel & Kerkhof, 1989). Increases in sleep time are more readily observed when recovery sleep is recorded during the dark portion of the light–dark cycle (Lancel & Kerkhof, 1989).

There was no significant difference in the number of Fos-IR or Fos + GAD-IR neurones in vlPOA during recovery sleep in SD rats versus SDC rats. However, a role for vlPOA neurones in homeostatic responses to sleep deprivation is supported by the finding that Fos + GAD-IR cell counts were significantly elevated in both SD and SDC rats compared to HSS rats (see Table 2). EEG delta power was elevated in SDC rats compared to HSS rats, indicating that SDC rats incurred at least a mild degree of sleep deprivation during the 24-h period on the treadmill. It is possible that maximal activation of vlPOA GABAergic neurones during recovery sleep occurs after only modest degrees of sleep restriction. In the rostral portion of the MnPN, the number of Fos + GAD-IR neurones during recovery sleep was similar in the SDC and HSS conditions. Severe sleep restriction may be required to increase c-fos expression in rostral MnPN neurones, and the mild amount of sleep loss incurred in the SDC condition was below the threshold of activation. Fos + GAD-IR cell counts in the caudal MnPN were elevated in SD versus SDC rats and in SDC rats compared to HSS rats. Of the three POA subregions examined, c-fos expression in GABAergic neurones located in the caudal portion of the MnPN appeared to be most responsive to varying degrees of sleep loss.

We also found that numbers of MnPN and vlPOA GAD-IR neurones that were also Fos-IR were significantly elevated in rats subjected to 26 h of sleep deprivation in the absence of any recovery sleep, compared to rats that underwent 24 h of SDC followed by 2 h of sleep deprivation. This finding suggests a role for activation of vlPOA and MnPN GABAergic neurones in mediating the increased sleepiness or sleep propensity during waking that occurs subsequent to sleep loss. Sherin et al. (1996) did not observe increases in Fos-IR cell counts in vlPOA in sleep deprived rats that were not permitted recovery sleep. However, the length of sleep deprivation was shorter than that described here (9–12 h versus 24 h) and the time of kill was different (7–9 h after lights on versus 3 h after lights-on).

The finding that Fos + GAD-IR cell counts in the MnPN are elevated during recovery sleep following SD as well as during waking following SD is consistent with our previous report that a subset of MnPN neurones exhibit progressive increases in activity during sustained waking, and during non-REM sleep that immediately follows episodes of sustained waking (Suntsova et al. 2002). We have previously shown that discharge of vlPOA neurones is elevated during recovery sleep following 12 h of sleep deprivation (Szymusiak et al. 1998). That finding is consistent with the increase in Fos + GAD-IR cell counts in SD and SDC rats compared to HSS rats in the present study. However, waking discharge rates of vlPOA neurones in sleep-deprived rats were similar to baseline values (Szymusiak et al. 1998), a finding that is not consistent with the observed increase in waking Fos + GAD-IR cell counts following SD in the present study. One possible explanation is that while waking Fos + GAD-IR cell counts were significantly elevated in SD versus SDC rats, the number of double-labelled cells was still substantially less than that observed during sleep. The ability to detect changes in waking discharge rates of only a subset of vlPOA neurones may require sampling a larger proportion of the vlPOA neuronal population than was done in our electrophysiological study (Szymusiak et al. 1998).

The intervention used to produce sleep deprivation was forced locomotion on a treadmill, with a 3 s on–12 s off duty cycle. This procedure resulted in near complete suppression of non-REM and REM sleep over a 24-h period. The method caused no overt behavioural signs of distress in SD rats, and we have previously shown that serum corticosterone levels in rats subjected to this SD procedure for 48 h were not elevated compared to levels in SDC and home cage control rats (Guzman-Marin et al. 2003). The SDC condition was designed to equate the total amount of forced locomotion incurred over a 24 h period with the SD condition, yet permit SDC rats the opportunity to achieve episodes of sustained sleep (duty cycle 15 min on–60 min off). The SDC condition also controlled for the environment of the treadmill, as rats were housed on the same apparatus under identical conditions during the SD and SDC procedures, with the exception of the treadmill duty cycle. Across the POA subregions examined, Fos + GAD-IR cell counts during recovery sleep differentially reflected the degree of sleep loss rather than non-specific factors associated with being placed on the treadmill. For example, Fos + GAD-IR cell counts in vlPOA were elevated in SDC versus HSS condition, with no further elevation in response to SD, while counts in the rostral MnPN were elevated in response to SD, with similar values in the SDC and HSS conditions.

Neuronal activation during sleep was inferred from the presence of c-Fos protein IR. While this is an indirect measure neuronal activity, a variety of stimuli that excite neurones have been shown to induce Fos protein (see Morgan & Curran, 1991 for review). Unit recordings in naturally sleeping animal corroborate Fos protein studies, demonstrating that neurones with elevated discharge rates during sleep compared to waking are prevalent in the MnPN and vlPOA (Szymusiak et al. 1998; Suntsova et al. 2002). However, direct evidence that preoptic area neurones that exhibit elevated discharge rates during sleep also express c-Fos protein during sleep is lacking.

Activation of neurones during sleep does not demonstrate a role for such neurones in sleep regulation. Sleep-related neuronal activation may reflect responses to autonomic or somatic changes that normally occur during sleep. However, additional evidence suggests a role for the preoptic area in sleep regulation, including the sleep suppressing effects of preoptic lesions (McGinty & Sterman, 1968; Szymusiak & Satinoff, 1984; Szymusiak et al. 1991; John & Kumar, 1998; Lu et al. 2000). Anatomical and physiological evidence supports the hypothesis that MnPN and vlPOA neurones function to promote and sustain sleep via descending inhibition of hypothalamic and brainstem arousal systems (McGinty & Szymusiak, 2000; Saper et al. 2001). The vlPOA is a major source of afferents to the histaminergic cell body region of the tuberomammillary nucleus in the posterior hypothalamus (Sherin et al. 1996, 1998) as well as to the serotonergic dorsal raphe nucleus, and the noradrenergic locus coeruleus (Sherin et al. 1996, 1998; Steininger et al. 2001). Projections of the MnPN include hypothalamic nuclei and associated structures involved in body fluid balance and neuroendocrine and visceromotor functions. These include the subfornical organ, supraoptic nucleus, the para- and periventricular nuclei of the hypothalamus, the dorsomedial hypothalamic nucleus, the posterior hypothalamic nucleus and the ventral tuberomammillary nucleus (Saper & Levisohn, 1983; Zardetto-Smith & Johnson, 1995; Thompson & Swanson, 2003). Projections to several putative arousal systems have also been identified, including the dorsal raphe and locus coeruleus (Zardetto-Smith & Johnson, 1995) and the perifornical lateral hypothalamus (Thompson & Swanson, 2003). Other studies have shown that descending projections to the posterior hypothalamus and rostral midbrain reticular formation from the POA and adjacent basal forebrain arise in GABAergic neurones (Gritti et al. 1994). MnPN and vlPOA sleep-active neurones exhibit a sleep–wake state dependent discharge pattern that is the reciprocal of the ‘REM-off’ discharge pattern observed in putative histaminergic, serotonergic and noradrenergic neurones (McGinty & Harper, 1976; Aston-Jones & Bloom, 1981; Vanni-Mercier et al. 1984; Szymusiak et al. 1998; Steininger et al. 1999; Jacobs & Fornal, 1999; Guzman-Marin et al. 2000; Suntsova et al. 2002) This finding is consistent with a possible functional inhibitory action of sleep-active neurones on the monoaminergic arousal systems. A subset of neurones recorded in the perifornical lateral hypothalamus also exhibits a ‘waking-related’ discharge pattern (Alam et al. 2002), which is reciprocal to the pattern observed in the majority of MnPN neurones. Activation of POA sleep-active neurones by local warming evokes suppression of waking discharge in the magnocellular basal forebrain (Alam et al. 1995b), the posterior hypothalamus (Krilowicz et al. 1994), the dorsal raphe nucleus (Guzman-Marin et al. 2000) and the perifornical lateral hypothalamus (Methippara et al. 2003). Collectively these findings suggest that GABAergic neurones in the MnPN and GABAergic/galaninergic neurones in the vlPOA exert sleep-related inhibitory control over these multiple arousal systems. The results of the present study further suggest that the strength of this inhibitory modulation is increased as a consequence of increasing prior time awake.

Acknowledgments

This work was supported by the Medical Research Service of the Departments of Veterans Affairs and NIH Grants MH63323, MH47480 and HL60296.

References

  1. Alam MN, Gong H, Alam T, Jaganath R, McGinty D, Szymusiak R. Sleep-waking discharge patterns of neurons recorded in the rat perifornical lateral hypothalamic area. J Physiol. 2002;538:619–631. doi: 10.1113/jphysiol.2001.012888. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Alam MN, McGinty D, Szymusiak R. Neuronal discharge of preoptic/anterior hypothalamic thermosensitive neurons: relation to NREM sleep. Am J Physiol. 1995a;269:R1240–R1249. doi: 10.1152/ajpregu.1995.269.5.R1240. [DOI] [PubMed] [Google Scholar]
  3. Alam MN, McGinty D, Szymusiak R. Thermosensitive neurons of the diagonal band in rats: relation to wakefulness and non-rapid eye movement sleep. Brain Res. 1997;752:81–89. doi: 10.1016/s0006-8993(96)01452-7. [DOI] [PubMed] [Google Scholar]
  4. Alam N, Szymusiak R, McGinty D. Local preoptic/anterior hypothalamic warming alters spontaneous and evoked neuronal activity in the magno-cellular basal forebrain. Brain Res. 1995b;696:221–230. doi: 10.1016/0006-8993(95)00884-s. [DOI] [PubMed] [Google Scholar]
  5. Aston-Jones G, Bloom FE. Activity of norepinepherine-containing locus coeruleus neurons in behaving rats anticipates fluctuations in the sleep-waking cycle. J Neurosci. 1981;1:876–886. doi: 10.1523/JNEUROSCI.01-08-00876.1981. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Findlay AR, Hayward JN. Spontaneous activity of single neurones in the hypothalamus of rabbits during sleep and waking. J Physiol. 1969;201:237–258. doi: 10.1113/jphysiol.1969.sp008753. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Gaus SE, Strecker RE, Tate BA, Parker RA, Saper CB. Ventrolateral preoptic nucleus contains sleep-active, galaninergic neurons in multiple mammalian species. Neuroscience. 2002;115:285–294. doi: 10.1016/s0306-4522(02)00308-1. [DOI] [PubMed] [Google Scholar]
  8. Gong H, Szymusiak R, King J, Steininger T, McGinty D. Sleep-related c-Fos protein expression in the preoptic hypothalamus: effects of ambient warming. Am J Physiol. 2000;279:R2079–R2088. doi: 10.1152/ajpregu.2000.279.6.R2079. [DOI] [PubMed] [Google Scholar]
  9. Gritti I, Mainville L, Jones BE. Projections of GABAergic and cholinergic basal forebrain and GABAergic preoptic-anterior hypothalamic neurons to the posterior lateral hypothalamus of the rat. J Comp Neurol. 1994;339:251–268. doi: 10.1002/cne.903390206. [DOI] [PubMed] [Google Scholar]
  10. Guzman-Marin R, Alam MN, Szymusiak R, Drucker-Colin R, McGinty D. Discharge modulation of rat dorsal raphe neurons during sleep and waking: effects of preoptic/basal forebrain warming. Brain Res. 2000;875:23–34. doi: 10.1016/s0006-8993(00)02561-0. [DOI] [PubMed] [Google Scholar]
  11. Guzman-Marin R, Stewart D, Suntsova N, Gong H, Szymusiak R, McGinty D. Sleep deprivation reduces proliferation of cells in the dentate gyrus of the hippocampus. J Physiol. 2003;549:563–571. doi: 10.1113/jphysiol.2003.041665. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Jacobs BL, Fornal CA. Activity of serotonergic neurons in behaving animals. Neuropsychopharmacol. 1999;21:9S–15S. doi: 10.1016/S0893-133X(99)00012-3. [DOI] [PubMed] [Google Scholar]
  13. John J, Kumar V. Effect of NMDA lesions of the medial preoptic neurons on sleep and other functions. Sleep. 1998;21:587–598. doi: 10.1093/sleep/21.6.587. [DOI] [PubMed] [Google Scholar]
  14. Kaitin KI. Preoptic area unit activity during sleep and wakefulness in the cat. Exp Neurol. 1984;83:347–357. doi: 10.1016/S0014-4886(84)90103-1. [DOI] [PubMed] [Google Scholar]
  15. Krilowicz BL, Szymusiak R, McGinty D. Regulation of posterior lateral hypothalamic arousal related neuronal discharge by preoptic anterior hypothalamic warming. Brain Res. 1994;668:30–38. doi: 10.1016/0006-8993(94)90507-x. [DOI] [PubMed] [Google Scholar]
  16. Lancel M, Kerkhof GA. Effects of repeated sleep deprivation in the dark or light period on sleep in rats. Physiol Behav. 1989;45:289–297. doi: 10.1016/0031-9384(89)90130-3. [DOI] [PubMed] [Google Scholar]
  17. Lu J, Bjorkum AA, Xu M, Gaus SE, Shiromani PJ, Saper CB. Selective activation of the extended ventrolateral preoptic nucleus during rapid eye movement sleep. J Neurosci. 2002;22:4568–4576. doi: 10.1523/JNEUROSCI.22-11-04568.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Lu J, Greco MA, Shiromani P, Saper CB. Effect of lesions of the ventrolateral preoptic nucleus on NREM and REM sleep. J Neurosci. 2000;20:3830–3842. doi: 10.1523/JNEUROSCI.20-10-03830.2000. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. McGinty D, Harper RM. Dorsal raphe neurons: depression of firing during sleep in cats. Brain Res. 1976;101:569–575. doi: 10.1016/0006-8993(76)90480-7. [DOI] [PubMed] [Google Scholar]
  20. McGinty D, Sterman MB. Sleep suppression after basal forebrain lesions in the cat. Science. 1968;160:1253–1255. doi: 10.1126/science.160.3833.1253. [DOI] [PubMed] [Google Scholar]
  21. McGinty D, Szymusiak R. The sleep-wake switch: a neuronal alarm clock. Nat Med. 2000;510:510–511. doi: 10.1038/74988. [DOI] [PubMed] [Google Scholar]
  22. Mendelson WB, Martin JV. Characterization of the hypnotic effects of triazolam microinjections into the medial preoptic area. Life Sci. 1992;50:1117–1128. doi: 10.1016/0024-3205(92)90349-t. [DOI] [PubMed] [Google Scholar]
  23. Methippara MM, Alam MN, Szymusiak R, McGinty D. Preoptic warming inhibits wake-active neurons in the perifornical lateral hypothalamus. Brain Res. 2003;960:167–173. doi: 10.1016/s0006-8993(02)03808-8. [DOI] [PubMed] [Google Scholar]
  24. Morgan JI, Curran T. Stimulus-transcription coupling in the nervous system: involvement of the inducible proto-oncogenes fos and jun. Ann Rev Neurosci. 1991;14:421–451. doi: 10.1146/annurev.ne.14.030191.002225. [DOI] [PubMed] [Google Scholar]
  25. Saper CB, Chou TC, Scammell TE. The sleep switch: hypothalamic control of sleep and wakefulness. Trends Neurosci. 2001;24:726–731. doi: 10.1016/s0166-2236(00)02002-6. [DOI] [PubMed] [Google Scholar]
  26. Saper CB, Levisohn D. Afferent connections of the median preoptic nucleus in the rat: anatomical evidence for a cardiovascular integrative mechanism in the anteroventral third ventricular (AV3V) region. Brain Res. 1983;288:21–31. doi: 10.1016/0006-8993(83)90078-1. [DOI] [PubMed] [Google Scholar]
  27. Scammell T, Gerashchenko D, Urade Y, Onoe H, Saper CB, Hayaishi O. Actviation of ventrolateral preoptic neurons by the smonogen prostaglandin D2. Proc Natl Acad Sci U S A. 1998;95:7754–7759. doi: 10.1073/pnas.95.13.7754. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Sherin JE, Elmquist JK, Torrealba F, Saper CB. Innervation of histaminergic tuberomammillary neurons by GABAergic and galaninergic neurons in the ventrolateral preoptic nucleus of the rat. J Neurosci. 1998;18:4705–4721. doi: 10.1523/JNEUROSCI.18-12-04705.1998. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Sherin JE, Shiromani PJ, McCarley RW, Saper CB. Activation of ventrolateral preoptic neurons during sleep. Science. 1996;271:216–219. doi: 10.1126/science.271.5246.216. [DOI] [PubMed] [Google Scholar]
  30. Steininger TL, Alam MN, Gong H, Szymusiak R, McGinty D. Sleep-waking discharge of neurons in the posterior lateral hypothalamus of the albino rat. Brain Res. 1999;840:138–147. doi: 10.1016/s0006-8993(99)01648-0. [DOI] [PubMed] [Google Scholar]
  31. Steininger TL, Gong H, McGinty D, Szymusiak R. Subregional organization of preoptic area/anterior hypothalamic projections to arousal-related monoaminergic cell groups. J Comp Neurol. 2001;429:638–653. [PubMed] [Google Scholar]
  32. Suntsova N, Szymusiak R, Alam MN, Guzman-Marin R, McGinty D. Sleep-waking discharge patterns of median preoptic nucleus neurons in rats. J Physiol. 2002;543:665–677. doi: 10.1113/jphysiol.2002.023085. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Szymusiak R, Alam N, Steininger TL, McGinty D. Sleep-waking discharge patterns of ventrolateral preoptic/anterior hypothalamic neurons in rats. Brain Res. 1998;803:178–188. doi: 10.1016/s0006-8993(98)00631-3. [DOI] [PubMed] [Google Scholar]
  34. Szymusiak R, Danowski J, McGinty D. Exposure to heat restores sleep in cats with preoptic/anterior hypothalamic cell loss. Brain Res. 1991;541:134–138. doi: 10.1016/0006-8993(91)91086-g. [DOI] [PubMed] [Google Scholar]
  35. Szymusiak R, Satinoff E. Ambient temperature-dependence of sleep disturbances produced by basal forebrain damage in rats. Brain Res Bull. 1984;12:295–305. doi: 10.1016/0361-9230(84)90057-1. [DOI] [PubMed] [Google Scholar]
  36. Thompson R, Swanson L. Structural characterization of a hypothalamic visceromotor pattern generator network. Brain Behav Rev. 2003;41:153–202. doi: 10.1016/s0165-0173(02)00232-1. [DOI] [PubMed] [Google Scholar]
  37. Ticho SR, Radulovacki M. Role of adenosine in sleep and temperature regulation in the preoptic area of rats. Pharmacol Biochem Behav. 1991;40:33–40. doi: 10.1016/0091-3057(91)90317-u. [DOI] [PubMed] [Google Scholar]
  38. Tobler I, Borbely AA. Sleep EEG in the rat as a function of prior waking. Electroenceph Clin Neurophysiol. 1986;64:74–76. doi: 10.1016/0013-4694(86)90044-1. [DOI] [PubMed] [Google Scholar]
  39. Ueno R, Ishikawa Y, Nakayama T, Hayaishi O. Prostaglandin D2 induces sleep when microinjected into the preoptic area of conscious rats. Biochem Biophys Res Commun. 1982;109:576–582. doi: 10.1016/0006-291x(82)91760-0. [DOI] [PubMed] [Google Scholar]
  40. Vanni-Mercier G, Sakai K, Jouvet M. Neurons specifiques de l'eveil dans l'hypothalamus posterior du chat. C R Acad Sci Paris. 1984;298:195–200. [PubMed] [Google Scholar]
  41. Zardetto-Smith A, Johnson A. Chemical topography of efferent projections from the median preoptic nucleus to the pontine monoaminergic cell groups in the rat. Neurosci Lett. 1995;199:215–219. doi: 10.1016/0304-3940(95)12003-m. [DOI] [PubMed] [Google Scholar]

Articles from The Journal of Physiology are provided here courtesy of The Physiological Society

RESOURCES