Korean Journal of Biological Psychiatry

Address for correspondence: Won Sub Kang, MD, PhD, Department of Psychiatry, College of Medicine, Kyung Hee University, 23 Kyungheedae-ro, Dongdaemun-gu, Seoul 02453, Korea
Tel: +82-2-958-8551, Fax: +82-2-957-1997, E-mail: menuhinwskang@khu.ac.kr

Introduction

ㅔSleep is a complex and intricate physiological and behavioral process. It can be defined by a distinct change in the electroencephalogram (EEG) from the waking state accompanied by behavioral change. Sleep is organized in two different states, rapid eye movement and non-rapid eye movement, which have different biological manifestations. Although unambiguous in its meaning, the function and exact pathophysiology of sleep are currently unknown.
ㅔThere are several hypotheses to explain the crucial role of sleep. For instance, sleep may be necessary for brain protein synthesis1) or to promote memory consolidation.2) Another possibility is that sleep promotes physical restoration. The Energy Allocation Model of Sleep3) suggests that limited energy resources are optimally distributed during sleep to essential biologic processes and decrease the total energy expenditure. After sleep deprivation in healthy control, 24 h energy expenditure increased by 7% compared to baseline.4) Recent studies support sleep enhances the early stage of the immune response. After 24 h sleep deprivation, monocyte production of interleukin-6 (IL-6) and tumor necrosis factor-alpha (TNF-α) increased,5) suggesting an association between sleep loss and inflammatory cytokine activity.
ㅔAlong with an effort to understand the function of sleep, many studies explored sleep's physiological and biochemical mechanisms. Neurons in the suprachiasmatic nucleus express a set of clock transcription factors that comprise the circadian feedback loop to create a daily oscillating cycle.6) In addition to the circadian system, multiple chemical mechanisms run the homeostatic systems. Adenosine accumulates during wake time to activate the inhibitory A1 receptor to enhance wake neurotransmitter.7)8) Among multiple neurotransmitters involved in the homeostatic sleep-pressure system, the importance of glutamatergic and GABAergic networks has been emphasized in recent years. GABAergic system targets the supramammillary nucleus, lateral hypothalamus, and the basal forebrain to promote wake.9)
ㅔIdentifying the expression changes of the genes between sleep and wakefulness may help clarify the core function of sleep at the cellular level. Previous studies have completed microarrays of RNA transcripts during sleep deprivation to identify functional categories of genes. Several gene expressions, including somatostatin,10) IL-1β,11) and those related to glycogen metabolism,12) peptidoglycan recognition protein,13) were identified as sleep state-dependent genes. After sleep deprivation, circadian genes per1 and per2 mRNA levels in the cerebral cortex increased while dbp expression decreased.14) Transcriptomics studies have concluded that gene expression in the brain changes considerably after sleep deprivation, and about 1% to 10% of gene expression showed changes in the array studies.15)16) However, most studies are done on rats and flies, and expression profiling is expected to be different in humans.
ㅔTo address the changes in gene expression after sleep deprivation in humans, we report a comprehensive microarray analysis of gene expression peripheral blood mononuclear cells (PBMC) to identify sleep state-dependent genes and highlight the functional consequence of sleep.

Methods

Subjects
ㅔAfter giving a written informed consent, eight healthy, normal male subjects aged 25–35 were included in the study. Before the study, an interview with a psychiatrist ensured that all participants had a regular sleep-wake rhythm. A battery of clinical tests, including vital signs, blood tests, and urine analysis were performed to ensure that subjects were physically and physiologically healthy. Exclusion criteria are the following: 1) abnormal values in blood test including basic chemistry, metabolic panel, and urine test, 2) comorbid mental health problems assessed by a psychiatrist, 3) use of drugs including alcohol, caffeine, psychotropic drug, and Z-drugs within ten days that could affect the results. The study was performed in October 2008. This study was carried out as recommended in the Helsinki Declaration. All participants signed an informed consent form, and the study received approval by the Ethical Review Committee (IRB No.: KMC GS-0808-03).

Protocol
ㅔEach subject participated in a sleep deprivation experiment that lasted for 24 h. A visual analog scale to assess appetite and mood and blood samples were done twice upon admission to Kyung Hee Medical Science Research Institute at 14:00 and after 24 h of sleep deprivation. The investigators were continuously present during the experiment to ensure that the subjects remained alert and would not fall asleep for 24 h of sleep loss. Kyung Hee Medical Science Research Institute is a hospital bed setting where subjects were allowed to lie down or engage in activities such as reading and playing video games. The subject was exposed to typical fluorescent lighting during the experiment.

Culture of human PBMC
ㅔObtained human peripheral venous blood was centrifuged (400 g) onto Histopaque (Sigma-Aldrich, St Louis, MO, USA). Cells were washed in phosphate-buffered saline, resuspended in RPMI 1640 medium, and cultured (5 × 10⁶ cells/mL) in 10-cm flat-bottomed culture plates, in medium supplemented with 25 mM HEPES, 10% fetal bovine serum, 2 mM l-glutamine, 100 U/mL penicillin, and 100 μg/mL streptomycin (WelGENE Inc., Seoul, Korea). After incubation at 37°C, 5% CO₂ air for 24 h, cells were collected in an Eppendorf tube and stored at -70°C.

cDNA synthesis
ㅔTotal RNAs of eight samples were extracted from PBM using the ZYMO research RNA isolation kit (Zymo Research, Irvine, CA, USA). Total RNA concentration and quality were accessed with the Agilent 2100 Bioanalyzer System (Agilent Technologies, Palo Alto, CA, USA). We used Agilent's Low RNA Input Linear Amplification kit PLUS to synthesize target cDNA probes and hybridization. T7 promoter primer mix and total RNA (1 μg) were incubated at 65°C for 10 min. cDNA master mix solution (5X first strand buffer, 0.1 M DTT, 10 mM dNTP mix, RNase-Out, and MMLV-RT) was mixed with the primer mixture and RNA. Reverse-transcription and synthesis of double-stranded cDNA (dsDNA) were finished by incubating the mixtures at 40°C for 2 h. After incubating the mixture at 65°C for 15 min, all the reaction was terminated. The transcription master mixture (4X transcription buffer, 0.1 M DTT, NTP mix, 50% PEG, RNase-Out, Inorganic pyrophosphatase, T7-RNA polymerase, and Cyanine 3/5-CTP) was added to the dsDNA reaction mixture. For dsDNA transcription, the mixture was incubated at 40°C for 2 h. Two cRNA samples before and after sleep deprivation were labeled with Cy3-CTP and Cy5-CTP, respectively. For fragmentation of cRNA, the mixture was incubated at 60°C for 30 min with 10X blocking agents and 25X fragmentation buffer. A 44 K Mouse Oligo Microarray (Agilent Technologies) kit containing 41174 transcripts and variants was used for microarray analysis. Scanned images in all groups were analyzed using Feature Extraction software (Agilent Technologies).

Microarray data analysis and statistical analysis
ㅔTranscriptional analysis of 41174 transcripts and variants of PBMCs was done before and after sleep deprivation. The ratio (CY5/CY3) of signal intensities generated in each microarray was normalized using Loewss models (Agilent's GeneSpring Software). Flag values were assigned to eliminate probes with a value lower than background signal intensity and achieve reliable gene expression profiles, and 38852 genes were identified as reliable genes. The signal intensity ratio (CY5/CY3) was computed and further log-transformed on the base of 2 into M = log2 (CY5/CY3) in order to treat the up-or down-regulation of identical magnitude as numerically equal but with opposite direction. Probes with a normalized average ratio greater than two in all eight participants were selected as significant genes. Statistical analysis was carried out using the program SPSS, release 12.0 (SPSS Inc., Chicago, IL, USA). Results are expressed as mean. Paired t-test was used to calculate statistical significance. Significance was determined if the p value was less than 0.5.

Results

ㅔWe tested for changes in gene expression with 24 h of sleep deprivation in healthy subjects. Among 41174 transcripts, 38852 genes were selected as reliable. We selected upregulated or downregulated genes with more than twofold changes after sleep deprivation because gene expression changes in peripheral blood are generally minor.17) Table 1 summarizes six gene expressions that were upregulated after sleep deprivation. Among the upregulated genes, protein tyrosine phosphatase receptor type O (PTPRO) was most upregulated (6.9-fold, p = 2.16 × 10^-5). Sleep deprivation increased the expression of the Fc fragment of IgG receptor IIIa, and IIIB (FCGR3A and FCGR3B) more than three-fold in PBMCs, which are related to natural killer (NK) cell-mediated cytotoxicity. Genes that were downregulated more than two-fold after sleep deprivation are a total of eleven (Table 2). Of the down-regulated genes, low-density lipoprotein receptor-related protein 5-like protein (LRP5L) showed the most substantial inhibition (0.06-fold, p = 4.47 × 10^-4). Nuclear receptor subfamily 4, group A, members 1 and 3 (NR4A1, NR4A3) were downregulated by 0.21 and 0.22-fold (p = 1.30 × 10^-5, 7.48 × 1^0-5) respectively.

Discussion

ㅔSleep deprivation affects man's biochemical, physiological, and psychological aspects, and due to high prevalence in the general public, it has become a significant concern. Although numerous studies examine the effects of sleep deprivation in each aspect, including cognition, mood, immune, and more, the overall understanding of gene expressions and general effects on human biology are obscure. This study investigates the gene expression pattern after 24 h of sleep deprivation in PBMC of healthy human subjects. Due to rare accessibility to brain tissue, we identified gene expression in peripheral blood nuclear cells that share more than 80% of the transcriptome with brain tissues.18) Of the 41174 genes assessed, only a small minority (< 1%) of the genes were up-or down-regulated (changes ranging up to two-fold) before and after 24 h of sleep deprivation.
ㅔSleep deprivation resulted in an elevation in the expression of the following inflammatory genes: PTPRO, IL8RBP, FCGR3A, and FCGR3B. PTPRO is one of the receptor types of phosphor tyrosine phosphatases and plays critical roles in acute inflammation via nuclear factor kappa B (NF-kB) signaling pathway in T lymphocytes19) and macrophages.20) PTPRO expression changes involve various types of cancers, including lung cancer,21) breast cancer,22) hepatocellular carcinoma,23) esophageal squamous cell carcinoma24) as well as inflammatory response, particularly toll-like receptor 4 mediated inflammation.25)26)
ㅔSleep deprivation increases the secretion of pro-inflammatory cytokines as well as increased white blood cells, most notably neutrophils.27) Previous studies reported notable elevation in IL-6 concentration28)29)30) after sleep deprivation and its role in quantity and depth of sleep. In our study, IL-6 increased by 1.8 fold, while IL8RBP gene expression showed a 4.87-fold change. IL-8 and cytokines such as IL-2, IL-6, IL-15, and IL-18 are reported to increase non-rapid eye movement sleep.31) Increased blood IL-8 concentration is observed in sleep disorders including narcolepsy32) and obstructive sleep apnea.33) However, gene expression and protein levels lack correlation,34) and further studies are required to understand the pathophysiology between sleep deprivation and IL-8 gene expression.
ㅔCD16 (FCγRIII) is expressed on macrophages, NK cells, neutrophils and is involved in diverse immune responses, including antibody-dependent cell-mediated cytotoxicity.35) Wilder-Smith et al.36) reported a significant decrease of CD 16 after 10 h-partial sleep deprivation and then increase above the baseline after prolonged sleep deprivation. The lag in NK activity in response to sleep deprivation is also consistent with other studies. A similar result is expected if we extend the observation time.37)
ㅔDown-regulated genes can be predominantly grouped into two functional pathways, metabolic and inflammatory pathways. Our study indicates possible effects on translation in the downregulation of glutamate and purine metabolism, including phospholipase A1 Member A (PLA1A), phosphodiesterase 4D (PDE4D), glutamine-fructose-6-phosphate transaminase 2 (GFPT2). GFPT2 codes enzymes that catalyze the conversion of glutamine to glutamate.38)39) Glutamate is an essential primary excitatory neurotransmitter that affects the sleep-wake regulatory system. Glutamate activates pre-synaptic glutamate receptors, thus increasing GABAergic inhibitory tone.40) Also, glutamate secretion affects astrocytes' production of ATP, critical gliotransmitters in regulating sleep demand.41)
ㅔThe orphan nuclear receptor 4A subfamily, including Nur77 (NR4A1), Nurr1 (NR4A2), and NOR1 (NR4A3), is emerging as a key regulator of inflammatory response. NR4A1 in monocytes and macrophages inhibits NF-κB activation by TNFα and IL-1β.42) It is also involved in T cell regulation and CD81 T cells development.43)44) In cancer, NR4A1 exhibits pro-oncogenic activities via TCR-induced apoptosis in T cell hybridoma.45) NR4A receptors all engage in inhibition of the inflammatory responses through various regulatory mechanisms. After sleep deprivation, two NR4A genes, NR4A1 and NR4A3, were downregulated, suggesting disengagement of inflammation inhibition response.
ㅔPDE4D is a cAMP-specific phosphodiesterase widely expressed in the human brain.46) Studies have shown that PDE45 interacts with several signaling proteins, most notably RACK1, β-arrestin2, MAPK signaling proteins, and oxidative-stress kinase. A possible functional role for PDE4D involves β2-adrenergic signaling,47) immunity/inflammation,48) and depressive disorder and cognition.49) Also, there seems to be a significant association between single-nucleotide polymorphism within PDE4D gene expression and sleep duration.50) Downregulation of PDE4D was observed in patients with atrial fibrillation.51)
ㅔIn addition to the change in immune-related gene expression, 19 clock genes were upregulated. Sleep deprivation is associated with changes in the expression of circadian clock-related genes in the cerebral cortex and PBMCs.52)53) None of the clock genes had a normalized average ratio greater than two in all participants in this study. Interindividual gene expression variability can be explained by polymorphic variants of clock genes.54) Mechanisms for gene expression changes induced by sleep and sleep deprivation in PBMCs and other tissues remain unclear, and further studies are required. PBMCs best represent the current status of the immune system and the inflammatory response, while the correlation of expression in blood with expression in the brain is imperfect. Given this, it may be possible that the immune-related gene expression change is more noticeable than the clock-related gene expression change.
ㅔOur findings suggest that of the 41174 genes assessed, some were specifically related to immune functions via up-regulation and down-regulation of immune and inflammatory responses in the absence of 24-h sleep deprivation. The study limitations are as follows: 1) Investigators were continuously present to monitor the participants to keep them alert during the study periods, but we were unable to record continuous EEG recording to ensure that the participants did not sleep. 2) The genome-wide expression analysis results need to be confirmed with a more extensive study number. Further studies to group the studied genes in biologically meaningful pathways to better understand the pathophysiological mechanism of sleep deprivation. 3) Small study samples and poor statistical analysis limits generalizability and interpretation of the data, and further studies to replicate the data are required. 4) Confounding variables such as food intake and activity level are not fully controlled. 5) In an ideal study, because acute sleep deprivation can alter the circadian expression of genes, blood samples should be drawn multiple times to understand the course of gene expression changes fully.
ㅔThe strengths of this study lie in that the subject is human, and PBMC can represent genetic changes in different organs after sleep deprivation, while previous studies are mostly done with animals such as rats. Also, the large number of genes allowed us to explore different possible pathways involved in sleep deprivation.
ㅔIn conclusion, the RNA-based approach provides a comprehensive gene expression profiling in PMBCs of sleep deprivation. There is a clear interaction between sleep deprivation and immune cascades via modification of inflammatory meditators level. These data should motivate further investigations to define mechanisms that explain the associations between sleep disturbance and inflammation.