Plasticity-Related Gene Expression During Eszopiclone-Induced Sleep

Abstract

Study Objectives:

Experimental evidence suggests that restorative processes depend on synaptic plasticity changes in the brain during sleep. We used the expression of plasticity-related genes to assess synaptic plasticity changes during drug-induced sleep.

Methods:

We first characterized sleep induced by eszopiclone in mice during baseline conditions and during the recovery from sleep deprivation. We then compared the expression of 18 genes and two miRNAs critically involved in synaptic plasticity in these mice. Gene expression was assessed in the cerebral cortex and hippocampus by the TaqMan reverse transcription polymerase chain reaction and correlated with sleep parameters.

Results:

Eszopiclone reduced the latency to nonrapid eye movement (NREM) sleep and increased NREM sleep amounts. Eszopiclone had no effect on slow wave activity (SWA) during baseline conditions but reduced the SWA increase during recovery sleep (RS) after sleep deprivation. Gene expression analyses revealed three distinct patterns: (1) four genes had higher expression either in the cortex or hippocampus in the group of mice with increased amounts of wakefulness; (2) a large proportion of plasticity-related genes (7 out of 18 genes) had higher expression during RS in the cortex but not in the hippocampus; and (3) six genes and the two miRNAs showed no significant changes across conditions. Even at a relatively high dose (20 mg/kg), eszopiclone did not reduce the expression of plasticity-related genes during RS period in the cortex.

Conclusions:

These results indicate that gene expression associated with synaptic plasticity occurs in the cortex in the presence of a hypnotic medication.

Statement of Significance

It is thought that mechanisms of synaptic plasticity underlie motor and cognitive recovery occurring during sleep. We hypothesized that similar synaptic plasticity changes can also occur during drug-induced sleep. To test this hypothesis, we studied expression of plasticity-related genes during drug-induced sleep. The results indicate that gene expression associated with synaptic plasticity can occur in the brain in the presence of a hypnotic such as eszopiclone.

INTRODUCTION

Disturbed sleep can lead to a wide range of health problems such as cognitive impairment, depressed mood, and negative effects on cardiovascular, endocrine, and immune function.¹ Sleep-inducing agents are often used to obtain sufficient amounts of sleep to avoid such problems. The restorative processes that occur during sleep have been suggested to depend, to a large extent, on synaptic plasticity.² However, the question of whether synaptic plasticity processes can occur as efficiently during drug-induced sleep as during physiological sleep has received little attention. Sleep-dependent plasticity processes have been studied in the primary visual cortex in the ocular dominance plasticity model.³ In this study, three commonly prescribed hypnotics (trazodone, zaleplon, and eszopiclone) had strong effects on sleep electroencephalogram (EEG) activity, but only trazodone significantly interfered with sleep-dependent consolidation of cortical plasticity.³ In the present study, we selected eszopiclone as sleep-inducing agent. Since eszopiclone has strong effects on sleep EEG activity but does not interfere with sleep-dependent consolidation of cortical plasticity,³ we hypothesized that eszopiclone would not interfere with the expression of plasticity-related genes during recovery sleep (RS) following sleep deprivation.

Because synaptic plasticity changes are manifested by upregulation of a number of plasticity-related genes, the expression of these genes may provide an indicator of plasticity that occurs during sleep. In the present study, we assessed the effects of eszopiclone on the expression of 18 genes that are critically involved in synaptic plasticity mechanisms. We compared gene expression in two brain regions (cortex and hippocampus) during spontaneous sleep with gene expression during RS following sleep deprivation either in the presence or absence of eszopiclone treatment. Since expression of the majority of plasticity-related genes is reduced during sleep,^4–6 expression of these genes was analyzed by quantitative real-time polymerase chain reaction (RT-PCR). We also included two relevant miRNAs as a target in our study because genes involved in synaptic plasticity are controlled at the translational level by miRNAs and the RNA-induced silencing complex (RISC).^7–9

METHODS

Experimental Animals

All animal experiments were carried out in accordance with the Guide for the Care and Use of Laboratory Animals and were approved by the Animal Care and Use Committee at SRI International. Thirty-one male C57BL/6 mice were obtained from Charles River Laboratories. All mice were maintained at an ambient temperature of 20°C–22°C under a 12-hour light-dark cycle (lights on at 07:00 hours) with food and water ad libitum.

Surgery

Mice were surgically implanted at 10 weeks of age with EEG and electromyogram (EMG) electrodes for sleep/wake recording. To record the EEG, each mouse was implanted with four stainless steel miniature screw electrodes placed bilaterally into the skull over the frontal and parietal cortices. To record the EMG, two electrodes made of Teflon-coated, stainless steel–braided wires were inserted into nuchal muscles. All electrodes were attached to a seven-pin electrical socket (Pinnacle Technology), and the entire assembly was fixed onto the skull with dental acrylic. Following surgery, each mouse was allowed a minimum of 3 weeks’ recovery followed by 1 week of adaptation to the recording cable and chamber.

Dose-Finding Study

The doses of eszopiclone (Sepracor Inc.) chosen for the dose-finding study were based on the published behavioral effects of zopiclone in mice¹⁰ and the sleep-promoting effects of zopiclone in rats.^11–14 In four mice that were surgically implanted with EEG and EMG electrodes and allowed to recover as described above, we tested three doses of eszopiclone (5, 10, and 20 mg/kg) along with a vehicle control (saline). Eszopiclone was dissolved in saline by solubilization in a small amount of 1 N HCl.¹⁴ The pH of the drug solution was adjusted to approximately 7 with 1 N NaOH.¹⁴ Eszopiclone was administered intraperitoneally (i.p.) in a volume of 0.1 mL. Doses were randomized and counterbalanced between animals, for a total of four injections/animal. Mice were injected at Zeitgeber Time (ZT) 6, and EEG and EMG were recorded for the next 4–6 hours. Three days elapsed between injections. Since no adverse EEG activity was detected after any of the doses tested, we used the highest dose (20 mg/kg) for further experiments.

Experimental Treatment and Sleep Recordings

Experiments were conducted in the following four conditions: (1) physiological sleep control (saline-injected; n = 7); (2) eszopiclone treatment (n = 6); (3) RS following sleep deprivation (saline-injected; n = 6); and (4) eszopiclone treatment during RS (n = 8). Figure 1 presents the experimental design for the four groups. Eszopiclone (20 mg/kg) or vehicle (saline) was injected i.p. in a volume of 0.1 mL in all mice 6 hours after the light onset (ZT6). EEG and EMG were recorded continuously at 22°C for 24 hours of baseline, for 6 hours of sleep deprivation, and during 160 minutes after cessation of sleep deprivation to determine the response to sleep deprivation. For groups 3 and 4, sleep deprivation was initiated at light onset (ZT0) by disturbing cage bedding around the mouse, stroking the vibrissae using an artist’s brush, and toward the end of the sleep deprivation period, stroking the fur with the brush when slow waves became evident in the EEG. Digitized EEG (digitization rate 100 Hz) and EMG (digitization rate 200 Hz) were collected via the Embla 16-channel A10 hardware system and visualized and stored with the Somnologica Science software application. Unihemispheric, frontal-parietal EEG potentials were filtered at 0.3 and 35 Hz and stored in 10-second epochs on a personal computer. EMG data were high-pass filtered ( > 10 Hz) for ease of visualization. At the end of the treatment period (Figure 1), brains from all mice were rapidly removed, the cerebral cortex and hippocampus dissected, flash-frozen on dry ice, and stored at –80°C.

Figure 1.

Schematic presentation of the experimental design. Mice were injected with saline or eszopiclone 6 hours after the light onset (ZT6). In groups 3 and 4, mice were sleep deprived from ZT0 to ZT6 and then allowed a recovery sleep opportunity from ZT6 to approximately ZT9. Brain tissue samples were taken from all mice at 9 hours after light onset (ZT9). The black and white bars at the base of the schematic represent the dark and light phases of the light/dark cycle.

EEG/EMG Data Scoring

Arousal states (wakefulness, rapid eye movement [REM] sleep, nonrapid eye movement [NREM] sleep) were determined in 10-second epochs using Sleep Sign (Kissei Comtec Co.) by a person blind to the type of drug administered to the mice. Wakefulness was identified by the presence of desynchronized EEG and high EMG activity. NREM sleep consisted of high-amplitude slow waves together with a low EMG tone relative to waking. REM sleep was identified by the presence of desynchronized EEG coupled with low EMG relative to NREM sleep. The amount of time spent in wakefulness, NREM, and REM sleep was determined in 20-minute bins. After the EEG/EMG recordings were scored, the code was broken to reveal the identity of each mouse. The EEG power spectrum was distributed into four frequency bands, namely, the delta (1–4 Hz), theta (4–8 Hz), alpha (8–13 Hz), and beta frequency bands (13–30 Hz). Relative delta power (ratio to total power) was calculated in 20-minute bins using Sleep Sign.

Gene Expression Studies

Total RNA was isolated from homogenate of the entire cerebral cortex and hippocampus using TRIZOL Reagent (Invitrogen) and TissueLyser (Qiagen). RNA was then purified using the RNeasy MiniKit (Qiagen). The RNase-free DNase Set (Qiagen) was used during RNA purification to remove any genomic DNA contamination. The RNeasy MiniKit and DNase digestion steps were omitted in case of miRNA amplification. First-strand cDNA was prepared from the cerebral cortex and hippocampus from each of the four groups (27 individual cDNA syntheses) using the TaqMan^® Reverse Transcription kit (Applied Biosystems, Foster City, California). For each reaction, a target cDNA of interest and the reference cDNA (glyceraldehyde-3-phosphate dehydrogenase; G3pdh) were simultaneously PCR amplified in duplicate or triplicate in 96-well plates along with eight concentrations of a mouse cortex RNA standard.^15,16 Thus, relative expression levels were determined in the cortex and hippocampus for all genes. Primers and probes for the plasticity-related genes were chosen using Primer Express v3.0 Software (Perkin-Elmer Applied Biosystems, Foster City, California) and are provided in Table 1. Primers and probes for the following genes were based on previous publications: Egr1,¹⁵Egr3,¹⁵Fra2,¹⁵G3pdh,¹⁵Egr4,¹⁷Mmp9,¹⁸Tpa,¹⁹ and Bdnf.²⁰ TaqMan^® MicroRNA Reverse Transcription kit and miRNA amplification kits were used to assess the miR-132 and miR-134 expression.

Table 1.

Primer and Probe Sequences Used for Quantitative Real-Time PCR Analyses.

Gene Symbol	Amplicon (bp)	Forward (F) and reverse (R) primers and probe (P) sequence
Arc	68	F: GCAGGTGGGTGGCTCTGA
		R: TCTTGGCTGGCCCATTCA
		P: 6FAM5′-AATATTGGCTGTCCCAGATCCAGAACCAC-3′TAMRA
Egr1	66	F: AAGACACCCCCCCATGAAC
		R: GCGAGAAAAGCGGCGAT
		P: 6FAM5′-CCCATATGCTTGCCCTGTCGAGTCC-3′TAMRA
Grp78	59	F: CACGTCCAACCCCGAGAA
		R: ATTCCAAGTGCGTCCGATG
		P: 5′FAM-CGGTCTTCGATGCCAAGCGCC-3′TAMRA
Grp94	64	F: CTTCCTTGTAGCAGATAAGGTCATTG
		P: 5′FAM-CACATCGAAACACAACAATGATACCCAGCA-3′TAMRA
		R: TCATTGGAGTCTGATTCCCAGA
Nr4a3	65	F: CCTCCGATCTGTATGATGAATGC
		R: GTGGGACAGTATCTGGAATAATCAAG
		P: 5′FAM-CGAGCTTTAACAGATGCAACACCCAGAGA-3′TAMRA
Fra2	90	F: TCTTCCTCAGCAGGGATGGA
		R: CAGAGGCTCTTCCCCGTAGA
		P: 6FAM5′-CCAGCGCTCTGTCATCAAGCCCAT-3′TAMRA
Egr3	68	F: TTGCTAAATCAATTGCCTGACAA
		R: CCGGAGAAGAGGTTGAGCG
		P: 6FAM5′-CTGTACCCCGAGGAGATCCCCAGC-3′TAMRA
Egr2	111	F: GGACCCAGGTCTCATTCCTATG
		R: GGACAGGGAAACGGCTTTC
		P: 6FAM5′-ATCCTGGATTTTTTCCATCTCCGTGCC-3′TAMRA
Egr4	79	F: TACAGCGGCAGCTTCTTCATC
		R: TGCCAGACATGAGGTTGAAGAG
		P: 6FAM5′-TTCCCGAACACCCGCACGACC-3′TAMRA
Mmp9	62	F: ACCCGAAGCGGACATTGTC
		R: CGAAGGGATACCCGTCTCC
		P: 6FAM5′-TCCAGTTTGGTGTCGCGG-3′TAMRA
Tpa	93	F: CAACAGCGGCCTGGTACAAT
		R: TACAGGGCCTGCTGACACGT
		P: 6FAM5′-TGCCTGTCCGAAGTTGCAGCGAAC-3′TAMRA
Camk2	88	F: TCCTGCCGCATTGAAGGA
		R: AACCAGCAGCCACATTCCA
		P: 5′FAM-CTTGCTTCGCAGAGATCCGCTCTTTG-TAMRA-3′
Homer	120	F: TTGACCCGAACACAAAGAAGAA
		R: TATTGCCTTTGAGCCATCTAAACTG
		P: 5′FAM-CCCACCAGCAAGCATGCAGTTACTGTATC-TAMRA-3′
Narp	134	F: GGCAAGCCAACGAGATTGTG
		R: AGTGGTCCAGGTGATGCAGAT
		P: 5′FAM-CTGCTCATCAACGACAAGGTCGCACA-TAMRA-3′
Neurogranin	91	F: GGCCAGAGCTGAACGTTTTAGA
		R: GCTCACAAACACAGTAGGGAAGTC
		P: 5′FAM-CGCGTCCCCTTCGCAGTGACA-TAMRA-3′
Scg2	84	F: TTTAATGCCCAATTTCCCTTCTT
		R: CACTGTCAACAGGATAGAGATCAACA
		P: 5′FAM-CCCCCAAGTAAGCCCCCTACATTTCTCT-TAMRA-3′
Syt4	105	F: GACAGAGCACGCAGAAAACATG
		R: GTGAAGACGAGGCCAAAAGC
		P: 5′FAM-CTCCTATCACCACCAGCCGCGTG-TAMRA-3′
Bdnf	68	F: CCATAAGGACGCGGACTTGTAC
		R: GAGGAGGCTCCAAAGGCACTT
		P: 5′FAM-CTTCCCGGGTGATGCTCAGCAGT-TAMRA-3′
G3pdh	65	F: CAACGGGAAGCCCATCAC
		R: CGGCCTCACCCCATTTG
		P: 6VIC5′-ATCTTCCAGGAGCGAGACCCCACTAACA-TAMRA-3′
hsa-miR-132		Applied Biosystems, Catalog Number 4373143
mmu-miR-134		Applied Biosystems, Catalog Number 4373299

Statistical Analysis

Sleep data were analyzed using two-way analysis of variance (ANOVA) with day (treatment vs. baseline) as a repeated factor—to control for individual differences in sleep parameters under baseline conditions—and treatment as a grouping factor (Statview 5.0.1; SAS Institute, Cary, North Carolina, USA). One-way ANOVA with treatment as grouping factor was used to compare gene expression between groups. Significant effects identified by ANOVA were followed by t-test with Student-Newman-Keuls corrections to identify differences.

RESULTS

Sleep Physiology Studies

Eszopiclone (20 mg/kg, i.p.) reduced the latency to NREM sleep after injection compared to the saline injections (Figure 2). This result is in agreement with previous studies in which zopiclone reduced latency to NREM sleep following the dose of 2 or 4 mg/kg p.o. in rats¹¹ and the dose of 10 mg/kg i.p. in mice.²¹ The latency to NREM sleep after eszopiclone injection was similar to that occurring after 6 hours of sleep deprivation (Figure 2).

Figure 2.

Effects of eszopiclone on the latency to sleep onset in undisturbed mice and in mice that were subjected to 6-hour sleep deprivation. Eszopiclone, sleep deprivation, or combination of eszopiclone and sleep deprivation significantly reduced the latency to sleep onset. Data represent the mean ± SEM percentage of wake, REM sleep, and NREM sleep from the total recording time. Statistical comparisons were made among all groups; letters at the top of bars indicate significant differences (p < .05) compared to the treatment groups indicated (from left to right: a–d). NREM = nonrapid eye movement; REM = rapid eye movement; RS = recovery sleep after sleep deprivation; SEM = standard error of the mean.

A significant decline in wakefulness and increase in NREM sleep amounts occurred during the 160 minutes after the injection of 20 mg/kg of eszopiclone without a change in REM sleep amounts (Figure 3). Both NREM sleep amounts and REM sleep amounts were greater in the RS or RS + eszopiclone groups of mice than in the saline-treated mice (Figure 3).

Figure 3.

Effects of eszopiclone on wakefulness, REM sleep, and NREM sleep in undisturbed mice and in mice that were subjected to 6-hour sleep deprivation. Eszopiclone (20 mg/kg) or saline was injected ip at ZT6, and sleep was recorded 160 minutes following the injection. Saline-treated animals had significantly higher amounts of wakefulness than eszopiclone-treated mice or mice recorded during the recovery sleep period (RS) after sleep deprivation. Data represent the mean ± SEM percentage of wake, REM sleep, and NREM sleep from the total recording time. Statistical comparisons were made among all groups; letters at the top of bars indicate significant differences (p < .05) compared to the treatment groups indicated (from left to right: a–d). NREM = nonrapid eye movement; REM = rapid eye movement; RS = recovery sleep after sleep deprivation; SEM = standard error of the mean.

Eszopiclone did not significantly affect the frequency (Figure 4A) or duration (Figure 4B) of the NREM or REM sleep bouts. However, eszopiclone + RS reduced the duration of wakefulness bouts (Figure 4B) and shorter wake bouts were evident with eszopiclone alone as well as RS alone.

Figure 4.

Effects of eszopiclone on the number of wakefulness, REM sleep, and NREM sleep bouts and bout durations in undisturbed mice and in the mice that were subjected to 6-hour sleep deprivation. (A) Mice of the saline-treated group had significantly fewer wakefulness bouts than mice of the eszopiclone-treated group. (B) Mice of the saline-treated group had significantly longer bouts of wakefulness than the mice treated with eszopiclone during the recovery sleep period (eszopiclone + RS group). Statistical comparisons were made among all groups; letters at the top of bars indicate significant differences (p < .05) compared to the treatment groups indicated (from left to right: a–d). NREM = nonrapid eye movement; REM = rapid eye movement; RS = recovery sleep after sleep deprivation.

During the 6-hour period of sleep deprivation, mice in group 3 spent 95.1 ± 1.4% of the time awake and 4.9 ± 1.4% of the time in NREM sleep and mice in group 4 spent 94.5 ± 1.2% of awake and 5.5 ± 1.2% in NREM sleep. As expected, the relative power density in the EEG delta frequency band was significantly increased during RS after sleep deprivation (Figure 5). However, this increase was transient in the mice treated with eszopiclone, suggesting that eszopiclone inhibits EEG delta activity. Surprisingly, a transient increase in EEG delta power also occurred during first 20 minutes after sleep onset in mice treated with saline. It is possible that a stress associated with the saline injection resulted in a delayed sleep onset (more than 30 minutes; Figure 2), allowing homeostatic drive to accumulate during the post-injection period.

Figure 5.

NREM sleep SWA profile on the baseline day (day 1) and on the injection day (day 2) in mice. Zero minutes correspond to ZT6, the time when injections occurred on day 2. Results are shown as the mean ± SEM. *p < .05 versus the same group of mice on day 1. NREM = nonrapid eye movement; SEM = standard error of the mean; SWA = slow wave activity.

Gene Expression Studies

Expression of 18 plasticity-related genes was measured by TaqMan quantitative RT-PCR in both the cortex and hippocampus of mice from the four different treatment groups. The expression levels of these genes were classified as three distinct patterns: wake-associated, RS-associated, and state invariant. Genes in the wake-associated category showed the highest expression in either the cortex or hippocampus of the saline-injected mice (group 1) and included Arc, Egr1, Egr2, and Egr4 (Figure 6).

Figure 6.

Real-time polymerase chain reaction analysis of the expression of the plasticity-related genes in the cerebral cortex (upper panel) and hippocampus (lower panel) across the four experimental conditions for the wake-associated genes. G3pdh expression was used as an internal standard. Values are mean ± SEM. p values are based on ANOVA; letters at the top of bars indicate significant differences (p < .05) compared to the groups indicated (from left to right: a–d). Note the increased level of expression in the group of mice treated with saline (group 1). ANOVA = analysis of variance; RS = recovery sleep; SEM = standard error of the mean.

Genes in the RS-associated category had higher expression levels in group 3 or 4 than in group 1 or 2 (Figure 7) and included Bdnf, Egr3, Grp78, Grp94, Homer, Narp, and Scg2. The mice belonging to groups 3 and 4 were sacrificed during RS after sleep deprivation, whereas the mice belonging to groups 1 and 2 were not subjected to sleep deprivation. Thus, this expression pattern is characteristic of genes that have a high expression level during RS.

Figure 7.

Real-time polymerase chain reaction analysis of the expression of the plasticity-related genes in the cerebral cortex (upper panel) and hippocampus (lower panel) across the four experimental conditions for the sleep-associated genes. G3pdh expression was used as an internal standard. Values are mean ± SEM. p values are based on ANOVA; letters at the top of bars indicate significant differences (p < .05) compared to the groups indicated (from left to right: a–d). Increased levels of expression are seen during recovery sleep after sleep deprivation in the mice injected with saline (group 3) or eszopiclone (group 4). ANOVA = analysis of variance; RS = recovery sleep; SEM = standard error of the mean.

Genes designated as “state invariant” did not show any significant changes in expression between the treatment groups (Figure 8). This category included Camk2, Fra2, Mmp9, Neurogranin, Syt4, and Tpa.

Figure 8.

Real-time polymerase chain reaction analysis of the expression of the plasticity-related genes in the cerebral cortex (upper panel) and hippocampus (lower panel) across the four experimental conditions for the state-invariant genes. G3pdh expression was used as an internal standard. Values are mean ± SEM. p values are based on ANOVA; letters at the top of bars indicate significant differences (p < .05) compared to the groups indicated (from left to right: a–d). Except for Nr4a3, there are no statistically significant differences in gene expression. ANOVA = analysis of variance; RS = recovery sleep; SEM = standard error of the mean.

Although miRNA expression had a tendency to increase following the eszopiclone treatment, these differences did not reach statistical significance (Figure 9).

Figure 9.

Real-time polymerase chain reaction analysis of the expression of two miRNAs in the cerebral cortex and hippocampus across the four experimental conditions. Values are mean ± SEM. ANOVA did not reveal any statistically significant differences in miRNA expression between groups. ANOVA = analysis of variance; RS = recovery sleep; SEM = standard error of the mean.

DISCUSSION

The experimental design of the present study differed from previous sleep-related gene-expression studies. Whereas several previous studies focused on comparing wakefulness versus sleep^4–6; comparison of spontaneous sleep versus sleep deprivation-associated RS has received relatively little attention.^15,22 We chose to test gene expression during the RS period because it has been proposed that restorative processes in the brain occur most efficiently during this period. According to the “synaptic homeostasis hypothesis”, restorative processes in the brain depend on synaptic plasticity changes. This hypothesis states that plastic processes occurring during wakefulness result in a net increase in synaptic strength in many brain circuits.^2,23–25 The role of sleep is to downscale synaptic strength to a baseline level that is energetically sustainable, makes efficient use of gray matter space, and is beneficial for learning and memory.²

At the molecular level, synaptic potentiation during wakefulness is reflected by an increase in the expression of a large number of plasticity genes (Arc, Camk2, Bdnf, Egr1, Homer, Narp, etc.).^26,27 In the hippocampus, such molecular changes are usually associated with long-term potentiation (LTP).^27,28 Although the “synaptic homeostasis” hypothesis predicts an overall reduction in synaptic strength during sleep, synaptic strength is expected to increase in at least some neuronal circuits such as those that underlie the reactivation of memory traces that were acquired during the preceding periods of wakefulness.^29,30 Hypnotics may have an effect on sleep-dependent brain plasticity,^3,31 but their effect on the expression of plasticity–related genes has been little studied. Therefore, we compared gene expression during spontaneous sleep with gene expression during RS following sleep deprivation either in presence or absence of eszopiclone treatment.

Effects of Eszopiclone on Sleep

Eszopiclone is the active stereoisomer of zopiclone, so the effects of eszopiclone and zopiclone on sleep are expected to be similar. Earlier studies in mice²¹ and rats¹¹ demonstrated that the primary effect of zopiclone on sleep is to increase the amount of NREM sleep. In one of these studies, zopiclone at a dose of 20 mg/kg i.p. did not significantly increase absolute amounts of NREM sleep in rats but significantly increased the NREM sleep time/total recording time ratio calculated within 6 hours of post-injection period.¹² The effects of eszopiclone on REM sleep are less consistent. A reduction in the amounts of REM sleep after 20 or 100 mg/kg i.p. of zopiclone was observed in one study in rats,¹² but no such reduction in REM sleep amounts was seen in other studies in which doses up to 10 mg/kg were used.^11,21

Consistent with other published studies,^13,21,32 eszopiclone reduced the latency to NREM sleep, increased NREM sleep amounts, and did not change REM sleep amounts (Figures 2 and 3). Similar to another study in mice in which zopiclone at 2–10 mg/kg did not modify the mean duration of NREM or REM sleep bouts or their frequency,²¹ there was also no change in the duration or frequency of the NREM or REM sleep bouts in the present study (Figure 4, A and B). Eszopiclone had no effect on slow wave activity (SWA) when it was injected during baseline conditions, but it reduced the SWA increase during RS after sleep deprivation (Figure 5). In this regard, the eszopiclone effect on EEG is similar to the effects of most of the presently used hypnotic drugs that alter sleep EEG parameters despite effectively improving sleep quality from the patient’s self-reported point of view.³³ In healthy young men, a single dose of midazolam (15 mg) or zopiclone (7.5 mg) reduced SWA and enhanced spindle frequency activity (11.25 to 15.0 Hz), but the time course of these parameters across and within sleep cycles as well as their mutual relationship were little affected.³⁴ These investigators concluded that hypnotics acting as benzodiazepine-receptor agonists do not substantially interfere with the homeostatic aspect of sleep regulation.³⁴

Effects of Eszopiclone on Expression of Plasticity-Related Genes

We chose 18 genes that have been critically implicated in various aspects of synaptic plasticity.^35–47 The other common feature of all these genes is that their expression changes between wakefulness and sleep states.^6,15,48,49 We assessed expression of these genes in the cerebral cortex and hippocampus because these brain regions exhibit significant changes in the expression of plasticity-related genes during the sleep–wake cycle.^4,5,6,49,50 Analysis of the gene expression revealed three distinct patterns.

Wakefulness-Associated Genes

The first pattern observed is represented by the highest gene expression levels in the mice treated with saline (group 1). We identified four genes with such a pattern: Arc, Egr1, Egr2, and Egr4 (Figure 6). Mice of group 1 had significantly higher amounts of wakefulness during the period of 160 minutes before sacrifice than mice of other three groups, suggesting that the amounts of wakefulness is the major factor affecting expression of Arc, Egr1, Egr2, and Egr4 in the cerebral cortex and hippocampus. This conclusion is supported by a number of other studies in which expression of the Arc and Egr family of genes markedly increased after extended periods of wakefulness.^6,15,27 In those studies, increases in Arc and Egr1 expression were among the most robust gene expression changes observed during wakefulness.^6,15 In our study, we observed a less robust increase in the level of expression of these genes (Figure 6), perhaps because the amounts of wakefulness were lower in the present study than in other studies. The maximum amount of wakefulness in the group 1 mice in our study was 43% of the total recording time, whereas the amounts of wakefulness were close to 100% in other studies in which sleep deprivation was performed. A recent study demonstrated that Arc, Egr1, and Egr2 belong to the group of “fast response” genes that are upregulated by sleep deprivation and remain elevated during approximately 2 hours of RS before returning to the original level.²² Since we sampled brain tissue after 2.5 to 3 hours of RS, this transient increase in expression at the beginning of the RS period was not detected.

The activity-regulated cytoskeletal-associated (ARC) protein is a key protein implicated in synaptic plasticity and memory consolidation.³⁵Arc mRNA is quickly induced and dynamically upregulated by behavioral experience.^35,51,52 It is thought that ARC and CAMK2 act as plasticity partners to promote functional and/or structural synaptic modifications.³⁶ The early growth response (EGR) family of transcription regulatory factors is implicated in orchestrating the changes in gene expression that underlie neuronal plasticity.³⁷ The most extensively studied member of this family is EGR1. EGR1 has been linked to the induction of hippocampal LTP,^53,54 neuronal morphological changes after exposure to an enriched environment,⁵⁵ and other plasticity-related phenomena.^56,57 The expression of other genes encoding EGR family members is also extremely sensitive to the environmental stimuli that cause plasticity.^37,58

RS-Associated Genes

The second pattern of gene expression consists of a higher expression in mice subjected to sleep deprivation with subsequent RS (group 3) or mice subjected to a combination of the eszopiclone treatment and RS (group 4) than in mice of the other two groups (groups 1 and 2). Remarkably, almost 40% of plasticity-related genes assessed in the present study showed this pattern of expression (Figure 7). This result is consistent with those of a recent study in which many genes including some of those examined in the present experiment (eg, Bdnf, Homer) were found to be upregulated during RS²². These results suggest that synaptic plasticity processes actively occur during the RS periods after sleep deprivation. However, it is also possible that this expression pattern is related to other cellular processes that require transcriptional regulation during first hours of homeostatic discharge of RS. Previous studies may not have detected such expression changes because those studies were conducted during spontaneous sleep associated with low homeostatic pressure (typically, after 6–12 hours of sleep).

Seven genes had increased expression in the cortex during RS either with or without eszopiclone treatment. These genes are Bdnf, Egr3, Grp78, Grp94, Homer, Narp, and Scg2. High expression of these genes is typically found after periods of extended wakefulness.^{15,16,59–63} We previously reported higher expression of Egr3, Grp78, and Grp94 in the cortex during the RS period than during the period of spontaneous sleep.^16,49 We now show that a large percentage of plasticity-related genes are highly expressed during the RS period. Despite such similarities in gene expression, the EEG patterns are very different between these behavioral states. High frequency (gamma or beta) predominate the EEG pattern during wakefulness.⁶⁴ Large increases in SWA are seen during RS, and SWA is especially high in the cortical areas that were actively involved in the learning during preceding wakefulness.⁶⁵ SWA was suppressed by eszopiclone during RS in the present study, but this suppression did not have an effect on the expression of plasticity-related genes in the cerebral cortex. Indeed, the level of expression of these genes was as high in the eszopiclone + RS group as in the RS group and even reached statistically significant difference for some genes (Grp94, Homer, and Narp) in the eszopiclone + RS group but not in the RS group (Figure 7). Thus, our results indicate that eszopiclone does not interfere with the expression of plasticity-related genes in the mouse cortex, even at the relatively high dose used in the present study (20 mg/kg).

Compared to the seven genes that had increased expression in the cortex during RS either with or without eszopiclone treatment, there were only two genes with a similar pattern of expression (Grp78 and Narp) in the hippocampus (Figure 7). This difference between the cortex and hippocampus may be related to temporal differences in the gene expression profile in these brain structures. If coordinated cortical and hippocampal replay of the multicell firing sequences evoked by awake experience²⁹ is followed by a longer phase of gradual strengthening of cortical-cortical synapses for long-term memory storage,⁶⁶ we would likely detect only the plasticity processes associated with the later phase because the gene expression was studied as long as 160 minutes after the onset of RS period in the present study.

Genes of the second expression pattern have been implicated in various aspects of synaptic function. BDNF is known to play a key role in the survival, growth, and maintenance of neurons during development⁶⁷ and to modulate synaptic plasticity in the adult brain.³⁸ Brain-derived neurotrophic factor (BDNF) affects synaptic plasticity through (1) the regulation of axonal and dendritic branching and remodeling⁶⁸; (2) synaptogenesis in arborizing axon terminals⁶⁹; (3) increasing the efficacy of synaptic transmission⁷⁰; and (4) the functional maturation of excitatory and inhibitory synapses.⁷¹ HOMER has been implicated in various models of synaptic plasticity.^72,73 The protein product of Homer is known to be targeted to the synapse where it interacts directly with other proteins in the postsynaptic density to facilitate changes in synaptic structure and function.^39,74Narp is a member of the pentraxin family.⁷⁵ Neuronal-activity-regulated pentraxin protein localizes specifically to excitatory synapses in primary neuronal cultures and in adult brain and is present in both the presynaptic and postsynaptic compartments.⁷⁶ It is an essential synaptogenic factor that is specific for excitatory synapses.^40,76Grp78 and Grp94 are frequently upregulated under conditions leading to plastic changes. They are classified as immediate early genes or molecular chaperones. For example, the chaperone function of Grp78 might serve to fold proteins and assemble protein complexes necessary for the structural changes characteristic of long-term memory.⁴¹ Chromogranins (A, B, and C) are secretory glycoproteins that coexist with peptide transmitters in large dense-core synaptic vesicles of many types of neurons.⁴² Chromogranin C (a product of the Scg2 gene) is released upon depolarization and is sensitive to changes in neuronal activity.⁴² Recent studies performed in Egr3-deficient mice suggest that Egr3 has an essential role in regulating cortical arousal, wakefulness, and sleep, presumably by its regulation of 5-HT₂ receptors.⁷⁷

State-Invariant Genes

The third pattern of gene expression observed was the absence of changes associated with either the eszopiclone treatment or RS. The genes with this expression pattern did not display statistically significant changes across four groups of mice by ANOVA analysis (Figure 8). Although Nr4a3 was significantly higher in the hippocampus of mice treated with eszopiclone (group 2) than in the mice of RS groups (groups 3 and 4), it was not presented here as a separate expression pattern because this pattern was unique to this gene (Figure 8).

Six genes did not show statistically significant changes in expression: Camk2, Fra2, Mmp9, Neurogranin, Syt4, and Tpa. The expression of these genes was found to differ between wakefulness and sleep in previous studies.^15,48,78 This discrepancy may be due to differences in the experimental design between the previous and present studies. In the present study, gene expression levels were determined about 160 minutes after the end of sleep deprivation. If sleep deprivation induces expression of these genes, a sufficient period of time may have elapsed for these genes to return to control levels. Future work may determine the time course of gene expression to establish the effects of wakefulness on expression of these genes.

Genes in the state-invariant category have been shown to play a role in synaptic function. CAMK2 is the main protein of the postsynaptic density.⁴³ It translocates to synapses and binds directly to the N-methyl-D-aspartate (NMDA) receptor and might act as a bistable switch for the long-term storage of synaptic memory.⁴³ A number of knockout, transgenic overexpression, and inhibitor studies have demonstrated that CAMK2 is essential for synaptic plasticity and learning.⁴³ Matrix metalloproteinases (MMPs) are a large family of endopeptidases, the substrates of which are proteins of the extracellular matrix as well as some nonmatrix proteins such as adhesion proteins.⁷⁹ MMPs act critically in regulating functional and structural remodeling of cellular architecture.⁸⁰ MMP9 plays a role for in hippocampal synaptic physiology, plasticity, and memory.⁴⁴ MMP9 protein levels and proteolytic activity are rapidly increased by stimuli that induce late-phase LTP in area CA1, whereas blockade of MMP9 pharmacologically selectively prevents induction of L-LTP.⁴⁴ A number of studies support the role of neurogranin in synaptic plasticity. Injection of antibodies to neurogranin into hippocampal CA1 pyramidal cells prevented induction of LTP in these neurons.⁸¹ In agreement with this finding, knockout of neurogranin resulted in a large decrease in the LTP induced by a single 100 Hz, 1s tetanus, whereas LTD was slightly enhanced.⁴⁵ Inducing LTP requires calmodulin stored in spines in the form of rapidly dissociating calmodulin–neurogranin complexes.⁸² Synaptotagmin IV is an abundant membrane protein that is localized to synaptic vesicles.⁸³ In cultured neurons, synaptotagmin IV is present in the Golgi apparatus and in distal parts of growing neurites, where its expression increases in response to membrane depolarization.⁴⁶ Because of its activity-dependent expression and its subcellular localization, synaptotagmin IV has been considered as a marker of synaptic plasticity. tPA, a serine protease classically known for its profibrinolytic role in the vasculature, has been implicated in numerous aspects of the synaptic plasticity process.^47,84 tPA has been shown to be upregulated in the cerebellum of rats during learning of a complex motor skill,⁸⁵ and the absence of tPA in tPA^–/– mice results in a significant reduction in the rate and extent of learning.⁸⁶

Effects of Eszopiclone on Expression of miRNAs

Since genes involved in synaptic plasticity are controlled at the translational level by miRNAs and the RISC,^7–9 we also evaluated expression of two relevant miRNAs in our study. Dendritic spine volume is regulated by miR-134, which appears to function through the translational control of LimK1, a regulator of actin polymerization.⁸ miR-132 is regulated by cyclic AMP-response element binding protein, an important regulator of different forms of plasticity.^9,87 In addition, the activity of both of these miRNAs may be controlled by BDNF—another major regulator of synaptic plasticity—via direct or indirect mechanisms.^8,9,88 We did not observe statistically significant differences in the levels of miR-132 and miR-134 either in the cortex or hippocampus (Figure 9).

Some limitations of the present study need to be acknowledged. Although we selected genes based on their role in plasticity, these genes are also implicated in other cellular processes. In addition, we did not measure protein concentrations. A large number of studies suggest a role for post-transcriptional, translational, and degradation regulation contribute to local protein concentrations.⁸⁹ Therefore, at least some proteins may be regulated in a different temporal pattern from that observed for mRNA in the present study.

CONCLUSION

Increased expression of synaptic plasticity-related genes has been found during wakefulness in many previous studies. We observed that a large proportion of plasticity-related genes increased expression also during RS (7 out of 18 genes included in the study). The significance of this increase in gene expression is currently unknown, but we hypothesize that it is related to restorative processes in the brain that actively occur during RS after sleep deprivation. Eszopiclone did not interfere with the expression of plasticity-related genes during RS in the mouse cortex even at a relatively high dose (20 mg/kg). These results suggest that synaptic plasticity changes and restorative processes can occur in the cortex in the presence of eszopiclone.

FUNDING

Research supported by Sepracor Inc., NIH R21 NS092926 and NIH R01 HL059658.

DISCLOSURE STATEMENT

None declared.