Paradoxical Sleep Deprivation: neurochemical, hormonal and behavioral alterations. Evidence from 30 years of research

Sleep comprises approximately one-third of a person’s lifetime, but its impact on health and medical conditions remains partially unrecognized. The prevalence of sleep disorders is increasing in modern societies, with significant repercussions on people’s well-being. This article reviews past and current literature on the paradoxical sleep deprivation method as well as data on its consequences to animals, ranging from behavioral changes to alterations in the gene expression. More specifically, we highlight relevant experimental studies and our group’s contribution over the last three decades.


INTRODUCTION
Sleep occupies approximately one-third of a person's life, but its impact on health and medical conditions remains partially unrecognized. The prevalence of sleep disorders is increasing in modern societies, where constant exposure to artificial light and interactive activities, such as television and the internet, combine with social and economic pressures to shorten the time spent asleep.
Most sleep disorders, such as sleep apnea, periodic leg movements disturbance, restless legs syndrome, jet lag, and insomnia, lead to sleep deprivation. Sleep deprivation per se represents a common type of stress that can have harmful physiological consequences, possibly leading to death in experimental animals (Rechtschaffen et al. 1983). Chronic insomnia leads to persistent tiredness and frustration due to lack of energy. Respiratory pathological conditions and other medical condi-tions that affect sleep have potentially serious com tions that may reduce life expectancy.
Disturbances in the sleep-wake rhythm and deprivation are increasingly frequent due to even are progressively more common in modern life i developed countries (Spiegel et al. 1999). Man jects can be chronically sleep-deprived as a re their current lifestyle (Miró et al. 2002). Sleep d tion can lead to traffic accidents due to fatigue an ations in the biological rhythm. Indeed, 16% of ian interstate bus drivers reported sleepiness whil ing, with an average of eight naps/trip (De Mell 2000), and 38% of those drivers have been dia with obstructive sleep apnea (Santos et al. 2004 cently, a new legislation was approved that will new medical and psychological examinations Brazilian drivers. These examinations will be i Sleep has a cyclical or episodic phase that alternates with wakefulness. An important characteristic of sleep that differentiates it from most of other states of altered consciousness is that it is promptly reversible. Electrophysiological studies in the 1950s demonstrated that there are two main states of sleep: non-rapid eye movement (NREM) and rapid eye movement (REM). The lighter stages of NREM sleep (phase 1 and 2) appear first and often alternate with brief episodes of wakefulness before the deeper NREM sleep stages set in. NREM sleep, particularly its deeper stages (phase 3 and 4), predominate early in the night, while REM sleep occurs at approximately 90-minute intervals. There are usually four to six sleep cycles each night, and, as the night progresses, the REM episodes become longer, and NREM episodes become both shorter and lighter. NREM sleep is recognized by its low frequency and high amplitude, in addition to the presence of sleep spindles and hypotonia as registered on an electroencephalogram (EEG).
Research suggests that REM sleep enables the establishment of behavior patterns and emotional responses. During REM sleep, the information obtained during wakefulness appears to be reprocessed and integrated into existing neural templates, so that future responses can be modified or matured to reflect both the experience of the individual and the inherent potential (Jouvet 1998). The electrophysiological features of REM sleep include recording of a wide range of desynchronized EEG frequencies, loss of EMG activity and rapid eye movement. The EEG reflects the intense cerebral cortical activity that distinguishes REM from A common irregular sleep-wake schedule is as follows: restricting sleep during a working week, leading to an accrual of sleep debt, and repaying the debt by sleeping longer on weekends. During these nights, there is high sleep efficiency, short sleep latency, and increased duration of stages 3 and 4 NREM sleep. Waking later in the morning, however, causes a phase delay the next day, which is often followed by an early wake-up time at the start of the working week. The duration of sleep is thereby considerably shortened, and sleep deprivation begins again (Shneerson 2000).
The homeostatic sleep drive appears to control NREM rather than REM sleep. Sleep is normally entered through NREM rather than REM sleep in adults, and an increase in the homeostatic drive will increase the duration and depth of NREM sleep at the expense of REM sleep. NREM sleep provides time for restorative processes within the central nervous system (CNS) and other parts of the body. Overall, NREM sleep appears to be a phase in which energy is conserved and in which both the CNS and other systems are able to either recover from the activity of the previous episode of wakefulness or to prepare for the next episode. Currently, most sleep deprivation occurs in the REM phase, which occurs during the last half of the total sleep session. It is in this phase that most restorative processes take place.
There are several different approaches that may be taken to understand the functional properties of PS. A common procedure involves deprivation of PS in various organisms and subsequent observation of behavioral and physiological changes. Thus, most studies of sleep deprivation focus on PS deprivation (PSD). The present paper reviews various studies that have utilized the PSD strategy.

PSD METHODS
The PSD technique entails an arousal of experimental animals by external stimulation at the onset of each PS period. Most of the instrumental methods utilized to induce PSD in animals are modifications of a single method developed by Jouvet et al. (1964) for cats, known as the bituated to this situation, the animals can obtain slow wave sleep (SWS), whereas PS is restricted. The loss of muscle tone associated with the onset of PS results in the animals touching the water and awakening. Information regarding the efficiency of the technique indicates that the degree of PSD may vary considerably. The diameter of the platform relative to the size of the animal seems to be mainly responsible for this variability (Steiner and Ellman 1972, Mendelson et al. 1974, Hicks et al. 1977. This method has been criticized because it submits the animal to additional adverse stimuli that could induce several of the effects observed after PSD. To attenuate the effects of these intervening variables, the multiple platform method was developed. In this paradigm, one rat is placed inside a large water tank containing seven platforms, which eliminates the restriction of movement and allows the animal to ambulate (Van Hulzen and Coenen 1981). This method was further modified in an attempt to eliminate the social isolation experienced by the animal in both the flower pot and the multiple platform techniques (Nunes and Tufik 1994a). In this paradigm, 10 animals on 18 narrow or wide platforms (as a control for the deprivation environment) are used, thus avoiding both social isolation and movement restriction. This technique is procedurally simple and allows for a large number of animals to be sleep deprived simultaneously. In 2000, Suchecki and Tufik proposed the modified multiple platform method (MMPM) using socially stable groups (10 animals from one cage) instead of animals coming from several cages. The results of this work suggest that the stress-related variables of the MMPM can be attenuated in stable groups.
Regarding the control group, most attention has been focused upon the selection of control subjects. When an animal is placed on a small platform in the water tank, many events happen in addition to PSD. For example, the animal is exposed to a good deal of stress. Control groups usually consist of either cage-control animals or rats placed on a wide platform (14 cm in di-factors inherent to this technique. Although this may be appropriate to address most of the confo factors, it remains controversial (Landis 1996(Landis , M et al. 2004. Indeed, when exposed to the M the animals placed on the wide platforms exhibit a hypertrophy similar to that produced by plac on narrow platforms (Nunes and Tufik 1994a). I tion, a wide platform induces some degree of as reflected by augmented rebound during the ery period (Machado et al. 2004). In the rotatio method proposed by Rechtschaffen et al. (1983 specific concomitants of this technique are con by a "yoked" control animal, which is concu aroused with the experimental animal (Van Hulz Coenen 1979). In the disk method, whenever perimental rat begins to sleep or enters a specifi stage, the disk automatically rotates, which awak experimental rat and forces both rats to walk in posite direction to the disk's rotation to avoid t ter (Rechtschaffen et al. 1983). Another propose trol group for the deprivation environment consi animals lying on a grid placed inside the wate (Suchecki and Tufik 2000). In this situation, the a could lie down without falling into the water, al their tails might have done so. The data suggest th is a good control; however, sleep rebound was ob The animal is exposed to a similar environment w in theory, being submitted to PSD. It is important that only social stable animals were placed on th as described in the same paper.
Recently, we proposed a paradigm for chroni restriction in rats in which the animals are subm a schedule of 18h of sleep deprivation with a 6h window for 21 days (Machado et al. 2005). As a of this restriction, there was a complete suppres PS and loss of SWS, which were partially compe during the 6h sleep window. Of note, a major slee solidation was observed during the 6h sleep oppo reflected by an intense reduction of awakenings.

524
SERGIO TUFIK, MONICA L. ANDERSEN, LIA R.A. BITTENCOURT and MARCO T. DE MELLO and Coenen (1984), rebound sleep is determined both by recovery processes and circadian timing. The increases (above baseline) in sleep duration during the recovery from sleep deprivation are usually only a fraction of the sleep time that was lost. Because it is widely assumed that sleep is homeostatically regulated, many believe that lost sleep is compensated by "intense sleep" that satisfies the unfulfilled sleep need in a relatively short period (Rechtschaffen et al. 1999). In addition to an increased sleep duration, the latency to the first PS episode is shortened, and the time spent in PS is lengthened (Dement 1960, Mendelson et al. 1974, Morden et al. 1967. Sleep rebound is also observed in rats submitted to different stress modalities , Palma et al. 2000, Papale et al. 2005. For instance, immobilization led to a rebound of SWS and PS, whereas cold stress produced an exclusive rebound of SWS and foot shock promoted sustained alertness during the animal's resting period. Together, these observations indicate that different stimuli alter the sleep pattern in distinct manners (Palma et al. 2000).

EFFECTS OF PSD ON BRAIN NEUROTRANSMITTERS
The nature of PSD techniques has led to much discussion about its consequences (Vogel 1975, Ellman et al. 1978. Most of the consequences are believed to be controlled by central neurotransmitter systems. To that end, many neurotransmitters, such as catecholamines, acetylcholine, serotonin, and GABA, have been associated, either alone or in trans-synaptic relationships, with several behavioral alterations. These associations suggest that the behavioral alterations observed after PSD result from deprivation-induced alterations in brain neurotransmitter function. Evaluation of experiments that employed the water tank PSD technique indicated that this procedure alters some behaviors that can be induced by drugs that act on neurotransmitter systems. The neurochemical basis of the constellation of behavioral changes that follow PSD is not completely clear, but many observations suggest a major role for central dopamine (DA). curred with greater magnitude and with smaller doses of apomorphine in the PSD rats when compared with control animals, and have been used to study the responsiveness of DA receptors to agonist agents. The authors suggested that the findings led to the belief that PSD induced supersensitivity of brain DA receptors. Shortly thereafter, further experiments supported the hypothesis that the exacerbated aggressiveness elicited by apomorphine in PSD rats was due to a state of supersensitivity of post-synaptic dopaminergic receptors in the brain (Tufik 1981a). This study showed that, in addition to apomorphine, PSD intensified the effects of two other dopaminergic agonists (bromocriptine and piribedil). The second finding of this study was that co-administration of PSD and an injection of haloperidol 24h before apomorphine administration induced more aggressive behavior when compared to that of rats that were only subjected to PSD, indicating that apomorphine was significantly more active in rats submitted to the double treatment. Finally, administration of haloperidol 2h before apomorphine administration blocked the aggressive behavior in rats submitted either to PSD alone or to PSD and an injection of haloperidol 24h previously. This result further supports the hypothesis that the postsynaptic dopaminergic receptors are in a state of supersensitivity.
Alternatively, these data may indicate that presynaptic dopaminergic activity was augmented by the above-described treatment regimen. Ghosh et al. (1976) reported that striatal DA levels were increased after PSD, raising the hypothesis that the hyperresponsiveness to apomorphine after PSD could merely reflect an excess of agonistic substances at the post-synaptic receptors. Tufik (1981b) reported that such a response was not due to a modification in DA turnover or levels, since it was demonstrated that animals treated with L-DOPA and injected with apomorphine failed to show aggressive behavior. However, PSD rats pre-treated with α-methylpara-tyrosine continued to show an intense response to DA agonists, suggesting that presynaptic mechanisms are not of prime importance for this augmented PSD-Interaction between the DA and norepinephrine (NOR) systems, which was achieved through modulation of the DA drug regimen, induced aggressive behavior in PSD rats that were pretreated with adrenergic drugs. This result indicates that NOR had a slight inhibitory action on the aggressiveness elicited by dopaminomimetic agents in PSD rats (Troncone and Tufik 1991). The reduced level of aggression could be due to a small increase in DA release associated with a lack of secondary noradrenergic inhibition rather than to a lack of NOR activity. This situation would evidence a role for beta-adrenoceptors in the modulation of this behavior. This proposition is supported by the observation that a reduced number of β-adrenoceptors are present in the cortex of rats submitted to PSD (Mogilnicka et al. 1986). This tissue has been shown to be subsensitive to NOR, which is reflected by evaluation of in vitro cyclic-AMP synthesis (Troncone et al. 1986). Autoradiographic analysis shows a generalized but regionally heterogeneous reduction in the number of β receptors subtypes in the brain, as well as clear trends of down-and up-regulation of α 1 and α 2 sites, respectively (Hipólide et al. 1998).
In another set of experiments that evaluated PSD and DA, yawning behavior was used to assess the sensitivity of DA and acetylcholine receptors in rats submitted to PSD (Tufik et al. 1987). The results demonstrated that PSD significantly decreased the yawning response induced by presynaptic doses of apomorphine and small doses of physostigmine and pilocarpine, suggesting that PSD induces subsensitivity of presynaptic DA receptors and/or postsynaptic acetylcholine receptors.
A large body of evidence indicates that drug-induced yawning in rats is a behavioral consequence of dopaminergic autoreceptor stimulation, which results in decreased DA synthesis and release, impaired dopaminergic transmission, and consequent removal of the inhibition of cholinergic neurons exerted by DA neurons (Yamada and Furukawa 1981). Thus, yawning behavior seems to involve both dopaminergic inhibition and cholinergic activation. By evaluating animals after a recovery period of 24h, Neumann et al. (1990) observed that, while apomorphine-induced yawning was still sig-minergic system occurs more strongly and at an time than on cholinergic system.
With respect to the evidence mentioned these findings led to the hypothesis that PSD i an upregulation of dopaminergic receptors (Tufi 1978(Tufi , 1987. However, because apomorphine is a D 1 /D 2 agonist, it was not possible to determin those experiments which subtype of dopaminer ceptor is unregulated after PSD. Nunes et al. ( used autoradiographic analysis to address this qu They found that PSD increased D 2 but not D 1 tor binding in the brain, suggesting that the unreg D 2 receptors are responsible for the previously re changes in apomorphine-induced behaviors afte An increase in DA release in the PSD group greater dopaminergic-mediated effects due to up tion of DA receptors in the striatum of PSD rat pared to control group (Farooqui et al. 1996). Th ing further supports the early studies on beh effects of dopaminergic stimulants in PSD anima fik et al. 1978fik et al. , Clark et al. 1987, Asakura et al. However, the decrease in yawning behavior indu cholinergic agonists (Tufik et al. 1987) led to an tigation of cholinergic receptors in animals sub to PSD. Again, an autoradiographic study show PSD resulted in a generalized down regulation type muscarinic receptors in the brain of a rat, c orating the notion that pontine M 2 -type recepto participate, but do not necessarily play an integr in the development of PSD effects (Nunes et al. 1 Some evidence suggests that PSD enhance tonergic transmission. Bearing this in mind, dir amination of 5-HT receptor binding after PSD w stricted to early studies that used a non-selective [3H]5-HT in a limited number of brain areas. studies showed that binding to 5-HT 1a was not cantly altered by PSD in any of the brain areas ined. However, there was an overall trend towa creased binding in the target paradoxically sleep de group. The same phenomenon occurred with ser transporter (SERT) binding after PSD. The freque ligand binding to 5HT sites was generally decrea Orexins are a newly discovered class of neuropeptides that regulate feeding and vigilance. Initially recognized for their importance in appetite control, orexins (also called hypocretins) are also involved in regulating sleep, arousal, and cardiovascular function.
Loss of orexin appears to be the primary cause of narcolepsy (Kilduff and Peyron 2000, Sutcliffe and de Lecea 2002). Expression of orexins is restricted to a discrete region of the hypothalamus, but the terminal projections of orexin-producing cells are widely distributed throughout the brain. In the rat, orexin induces a dose-dependent increase in wakefulness when it is injected intracerebroventricularly (Piper et al. 2000). The densest projection of orexinergic cells in the rat is in the locus coeruleus (Horvath et al. 1999), the noradrenergic output that favors cortical arousal of waking and opposes PS-associated arousal (Hobson et al. 2000, Aston-Jones et al. 2001. Pedrazzoli et al. (2004) measured hypocretin levels at 6 and 96h after PSD and following 24h of rebound. PSD was found to increase cerebral spinal fluid orexin-1 collected at zeitgeber time (ZT) 8, but not at ZT0. Decreased cerebral spinal fluid orexin levels were also observed at ZT8 after 24h of PS rebound. These results suggest that PSD activates and that rebound inhibits the orexin system. The authors suggested that increased orexin tone during PSD might be important for some of the effects of PSD, such as antidepressant effects, hyperphagia, and increased sympathetic activity. Another experiment performed by D' Almeida et al. (2005) showed that the level of prepro-orexin (the precursor of both forms of orexin) mRNA increased by 30% at 96 hours after in the PSD group and by 88% after 24 hours in the rebound group relative to controls. Sleep deprivation produced a heterogeneous effect on brain mRNA levels of both orexin receptors (OXR). The OX 1 R mRNA levels increased in more than 30 different regions, particularly in the amygdala and hypothalamus in the rebound group compared to control and sleep deprived groups. Changes in OX 2 R mRNA levels were also seen only in the sleep rebound group, but and alters orexin receptor expression. These effects are not immediately reversed by sleep recovery, indicating a lasting activation of the orexinergic system. Collectively, the overall changes in brain neurotransmitters after PSD may indicate that the mechanisms involve in the process of sleep recovery.

PSD AND GENE EXPRESSION
Sleep deprivation of short duration or prolonged for several days alters the expression of distinct categories of genes in the brain, including immediate-early genes, such as transcription factors; genes that encode proteins involved in metabolic processes, neuronal plasticity, and the stress response; and genes that encode neurotransmitters, hormone receptors and transporters, and enzymes (Cirelli et al. 2004, Cirelli 2006, Terao et al. 2008, Maret et al. 2007). In addition, it has been shown that the pattern of changes is consistent across species (Cirelli 2002) and in distinct brain regions (Cirelli et al. 2004, Terao et al. 2008.
Through the use of microarrays to evaluate the expression of > 30, 000 transcripts, we recently identified a total of 78 unique transcripts that were differentially expressed in animals submitted to PSD relative to controls (Guindalini et al. 2009). These differentially expressed transcripts include genes from many functional classes, including metabolic processes, the circadian sleep-wake cycle, the response to a stimulus, regulation of cell proliferation, and signaling pathways. After 24h of sleep rebound, expression of ∼ 61.5% of the sleep deprivation-regulated transcripts was detected as being differentially expressed in the recovery period, resembling the gene expression pattern of the controls, relative to the PSD group. However, as occurs with a number of behavioral and physiological parameters, the expression levels of the remaining 38.5% of transcripts were among the levels observed for control and sleep-deprived animals, suggesting that a longer recovery period is necessary to completely reverse the effects of prolonged PSD. Noteworthy, DNA damage is an important step in events leading to genomic instabil-the liver or heart cells (Andersen et al. 2009). This study represents a relevant contribution to the understanding of the potential health risks associated with sleep deprivation. Collectively, these results suggest that PSD elicits a pattern of gene expression that is similar to those observed for other paradigms of sleep deprivation, and that PSD has characteristics that might be related to the regulation of PS. In addition, 24h of sleep recovery was not sufficient for complete reversal of the effects caused by prolonged periods of PSD of a number of behavioral and physiological parameters (Andersen et al. 2005a, Antunes et al. 2006. Interestingly, Maret and colleagues demonstrated that sleep deprivation induced a larger number of transcriptional changes in the liver than in the brain, suggesting that sleep deprivation might have a strong effect on peripheral tissues and/or that the brain might be protected from major changes in transcription (Maret et al. 2007). This study corroborates our results, which indicate that there is less variation in the gene expression in the brain compared to other tissues, such as blood and cavernosum, following PSD (Lee et al. 2009).

YAWNING
Apart from the constellation of behavioral and neuropharmacological consequences of PSD, yawning is particularly interesting because it can be elicited by several cholinergic agonists (Urbá-Holmgren et al. 1977, Wood et al. 1979), such as low doses of dopaminergic agonists (Mogilnicka and Klimek 1977) and polypeptides such as α-MSH and ACTH (Ferrari et al. 1963, Yamada and Furukawa 1981, Lobo et al. 1990). This finding suggests that multiple neurotransmitter systems modulate this behavior. PSD for 96h results in nearly complete suppression of yawning induced by dopaminergic and cholinergic agonists (Tufik et al. 1987). Since the activity of both of these neurotransmitter systems is also altered by stress (De Kloet 1991), animals were chronically exposed to different stress modalities. The findings suggest that yawning is differentially by constant and intermittent stressors (Tufik et al. As these stressful manipulations altered d duced yawning, Hipólide et al. (1999) investiga effects of single and repeated treatments with thetic glucocorticoid, dexamethasone (DEXA) o morphine-and pilocarpine-induced yawning i Neither single nor repeated treatment with DEX tered apomorphine-induced yawning. A single tion of DEXA, however, caused an increased n of yawns induced by pilocarpine. Repeated tre with DEXA led to a decreased number of yaw duced by pilocarpine. The authors concluded tha minergic and cholinergic systems are distinctly by DEXA in terms of yawning behavior. Furthe yawning behavior was evaluated to examine w concomitant treatment with PSD and DA agonist reverse PSD effects , as observe stereotypy and aggressiveness (Troncone et al. 19 More recently, the effects of acute and chron and 9 -tetrahydrocannabinol ( 8 and 9 -THC examined on yawning induced by pilocarpine o morphine. The data suggest that cannabinoid a inhibited yawning induced by cholinergic and do ergic agonists. In addition, the increased freque spontaneous yawning following cessation of chro ministration of a cannabinoid agonist may be of tance as a withdrawal sign for these drugs (Nak Palacios et al. 2002). Cannabis also induces agg behavior in PSD rats. This effect is probably rel brain catecholamines, as DA may play an agon and NOR an inhibitory role (Carlini et al. 1977). COGNITION Sleep has roles in learning and memory proce several levels (Stickgold and Walker 2007). Tw lines of evidence support this hypothesis in r Firstly, PS following the acquisition of learning leads to increases in the number and duration o odes and PS density (Smith and Wong 1991). Sec many studies report an impaired acquisition whe because the procedure affected inhibitory avoidance and an appetite-motivated task (Stern 1971). In contrast, using appetitive motivation, no effect on accuracy of Tmaze discrimination was found with PSD (Hicks and Paulus 1973). The nature of the task might be one factor that accounts for these discrepancies. To determine whether PSD administered prior to training would have differential effects on two aversively motivated tasks (classical conditioning of fear and inhibitory avoidance), Bueno et al. (1994) submitted rats to PSD, and the rats were trained on a double aversively motivated task. The data show that PSD administered 24 or 72h before the training session had no effect on either task. However, when the period of deprivation was extended to 96h before the training session, acquisition of inhibitory avoidance was impaired, but no effect on classically conditioned fear was observed. Thus, the authors concluded that PSD had differential effects on the two tasks, both aversively motivated and trained at the same time and under the same conditions. Moreover, the study on the effects of pharmacological interference with the cholinergic system on PSD-induced memory impairment show that activation of this system during the period of deprivation can prevent memory deficits induced by PSD (Bueno et al. 2000). Later data from our group suggest that proactive learning/memory deficits induced by PSD could be attributed to altered M1 binding either immediately after PSD (when avoidance training occurs) or after sleep recovery (when acquisition/retention are tested) (Moreira et al. 2003).
In 2000, Silva and Frussa-Filho proposed the use of the plus-maze discriminative avoidance task (PM-DAT), which is a behavioral paradigm that has the advantage of independently evaluate learning, memory, and anxiety as well as motor activity. PM-DAT complies with ethical protocols since there is no foot shock, which is imposed by classical avoidance tests. Considering the effects of PSD in mice submitted to the PM-DAT (Silva et al. 2004b), pre-training PSD-induced anxiogenic and hypolocomotor effects were observed in the training session, but they did not modify the acquisition of the task. neously evaluates learning, memory, anxiety and motor function. Our results demonstrate that PSD induced impairment in the acquisition, consolidation and retrieval of a discriminative avoidance task. In the three experimental conditions, a 24h sleep recovery period was effective in abolishing the PSD cognitive effects, but was ineffective in modifying them per se. In addition, PSD produced significant alterations in anxiety-like behavior and locomotor activity, which may be related to the impairment in retrieval (but not in acquisition or consolidation). This study strengthened the critical role of PS (but not sleep rebound) in all the phases of learning and memory formation. In addition, it suggests that PSD effects on acquisition and consolidation do not seem to be related to other behavioral alterations induced by this procedure.

LOCOMOTOR AND ANXIETY-LIKE BEHAVIOR
Regarding the effect of PSD on locomotor activity (Tufik and Silveira-Filho 1983, Frussa-Filho et al. 2004, Andersen et al. 2005b, Perry et al. 2007, some studies have focused on locomotor activity and sleep patterns. Consequently, speculations about the effect of the exercise on sleep due to sleep deprivation have been raised (Martins et al. 2001).
From our data, it can be assumed that PSD induced significant but heterogeneous effects in animals. For instance, PSD increased grooming, but it had no effect on stereotyped behaviors, locomotion, or the elevated plus maze test. Importantly, it significantly decreased rearing behavior, which is associated with potentiation of stereotypy and exploratory activity (Andersen et al. 2005b). In addition, another study showed that single platform PSD rats had reduced locomotor activity and augmented anxiety-like behavior (Suchecki et al. 2002). Recently, Martins and colleagues (2008) pointed out that PSD increased gnawing behavior undirected to feeding, supporting the view that spontaneous oral stereotypy, rather than hunger, is the major cause of increases in food removed from feeders in rats deprived of sleep for up to 96 hours. al. 2004a, b), as well as in the open-field test (decreased in central locomotion frequency). This anxiogenic effect was not observed after 24h of sleep deprivation.

SEXUAL BEHAVIOR
As mentioned, PSD results in several behavioral alterations in rats that are likewise induced by dopaminergic agonists (Tufik et al. 1978, 1987, Andersen and Tufik 2005; however, there is some controversy about the effect of PSD on sexual behavior. DA plays a key role in the modulation of this behavior (Bitran and Hull 1987). Morden and co-workers (1968) reported increased sexual performance in PSD male rats. In another study, however, four days of PSD did not significantly alter male sexual performance (Hicks et al. 1991).
In this context, we have consistently demonstrated the facilitatory effects of PSD on genital reflexes (penile erection, or PE, and ejaculation, or EJ) in rats (Andersen et al. 2000). The data revealed that only PSD rats exhibited PE and EJ behaviors, which were absent in control animals. Indeed, combined administration of PSD and cocaine elicited genital reflexes much more markedly than did administration of PSD or cocaine alone (Andersen and Tufik 2002). Although a number of factors are involved in a complex phenomenon such as PE, the authors suggest that the previously documented DA receptor supersensitivity induced by PSD may be an important contributor to the potentiation of genital reflexes by cocaine after sleep deprivation. Importantly, other dopaminergic drugs, such as cocaine (Andersen et al. 2004a, b), methamphetamine (Andersen et al. 2003a) and apomorphine (Andersen et al. 2003b), also increased the effects of PSD on genital reflexes in both young (Andersen et al. 2003c) and old male rats (Andersen et al. 2002(Andersen et al. , 2004a.
The period of exposure to PSD seems to interfere with the responses of genital reflexes. After 24h of PSD followed by cocaine administration, adult animals started to display PE and EJ, which peaked at 96h of PSD. More than 96h of PSD decreased genital reflexes, as observed at 120 and 144 hours (Andersen et level of testosterone, whereas progesterone level increased in both young and old groups, suggestin although sexual function commonly decreases wi testosterone alone cannot be considered the ma mone involved in sexual activity (Andersen et al. 2003c(Andersen et al. , 2004a.

ESTROUS CYCLE
Thus far, little attention has been given to whethe is differentially regulated between genders and th nitude of the consequences of sleep loss. Studie documented that, while healthy women appear t better sleep quality than men, they report more problems, including inadequate sleep time and nia (Bixler et al. 2002, Zhang andWing 2006). over, nightmares were reported to be twice as fr in women (Ohayon et al. 1997). The reasons fo discrepancies are attributed to hormonal fluctu over the menstrual and estrous cycles, a factor th been associated with sleep variations in both h and rats (Hachul et al. 2006, Antunes et al. 2006 In particular, alterations in reproductive ho release coincidental with sleep have been sugge a manifestation of entrained links between CNS tion and endocrine function. Notably, our results d strate that female rats submitted to PSD in the di phase (PSD-diestrous) had their estrous cycles dis during the recovery period. This was evidence constant diestrus during the first week of recove for hormone alterations, progesterone concent were statistically higher in PSD-diestrus compa the respective phase control and to PSD-pro and PSD-estrous rats, while CTRL-metestrous r higher levels than CTRL-proestrous and estrous g In addition, the PSD-diestrous phase exhibited concentrations of corticosterone and lower es levels than the respective control rats. These d dicate that PSD may modulate ovarian hormone re through alterations in hormonal-neurochemical sy OXIDATIVE STRESS  et al. 1989et al. , Everson et al. 1994, as well as motor weakness, lesions in organs, and even death (Rechtschaffen et al. 1983). It is possible that these effects occur as a result of reactive oxygen species (ROS) accumulation. A number of studies have reported that ROS are involved in the pathogenesis of several diseases and psychological stress (Halliwell and Gutteridge 1989). D'Almeida and coworkers have studied the oxidative stress status following PSD. First, they described a reduction in total glutathione (GSH) levels in the whole brain; however, a number of antioxidants were found to be unchanged in the brain after PSD (D'Almeida et al. 1997). Furthermore, reduced GSH levels were demonstrated in the hypothalamus and thalamus (D'Almeida et al. 1998), suggesting that these specific brain regions could be more susceptible to oxidative stress during PSD (D'Almeida et al. 2000). To investigate the hypotheses that (i) sleep involves a process of detoxification at the cellular level (Inoué et al. 1995) and (ii) that sleep may serve as an antioxidant function by removing free radicals or ROS produced during waking (Reimund 1994), and, considering that GSH levels were counteracted by melatonin administration (Floreani et al. 1997), the authors tested the effects of exogenous melatonin administration on the PSD-induced decrease in GSH. Twice daily administration of melatonin during the PSD period, and equivalent dosing in controls, resulted in similarly lowered hypothalamic GSH levels, indicating that the antioxidant properties of melatonin were not able to preserve hypothalamic GSH levels after PSD, and that exogenous melatonin may lead to decreases in hypothalamic GSH levels and further the decrease induced by PSD in this brain region (D'Almeida et al. 2000).
Reduced GSH levels, a key indicator of oxidative stress, have been shown to precipitate apoptotic cell death in the brain under different conditions (Tan et al. 1998). Based upon the above-described findings of D'Almeida and colleagues, a study was performed that showed that binding of the "peripheral-type" benzodiazepine ligand [ 3 H]PK 11195 to reactive astrocytes, a of 24 brain areas studied (Hipólide et al. 2002) were likewise unchanged. These findings suggest that the oxidative stress that develops in the brains of rats submitted to PSD does not result in cell loss.
Considering the stressful nature of the PSD method, which is associated with the disruption of various physiological processes, D'Almeida's group described that plasma homocysteine (an intermediate amino acid in methionine-cysteine metabolism) was reduced in PSD rats as compared with the cage group, and did not revert to normal levels after 24 or 48h of recovery. Decreased glutathione and increased thiobarbituric acid reactive substance levels were observed in PSD rats (Oliveira et al. 2002). The authors suggested that the observed decreases in homocysteine levels may represent a self-correcting response to depleted glutathione in PSD rats, which would contribute to the attenuation of the deleterious effects of PSD. Homocysteine levels were decreased in older rats as compared to young ones; PSD did not further reduce the lowered levels in aged animals. Additionally, PSD had no effect on total cholesterol or folate levels, while it increased HDL, LDL, and vitamin B 12 , and decreased triglyceride and VLDL levels (Andersen et al. 2004c). Indeed, the increase in LDL levels, which was much more pronounced than the increase in HDL, would constitute an unfavorable factor since the LDL/ HDL ratio represents a better assessment of cardiovascular disease risk than HDL cholesterol levels alone (Lemieux et al. 2001). These results indicate that PSD has significant but heterogeneous physiological effects in aged rats, and it may intensify certain age-related effects that contribute to cardiovascular disease risk (Andersen et al. 2004c).
One of the most common sleep disorders, obstructive sleep apnea, is characterized by an airflow interruption despite persistent respiratory efforts, thus causing chronic sleep deprivation due to the frequent awakening at the termination of apneic episodes. As sleep deprivation may aggravate hypertensive features, and since cardiorespiratory changes are observed after PSD, Palma et al. (2002) studied the participation of endothelin-1 in trations, suggesting that the increase in the endothelin levels may be involved in the genesis of arterial hypertension and cardiorespiratory changes, which are observed after sleep deprivation (Palma et al. 2002).
It has been proposed that both hypoxia and sleep fragmentation are implicated in the cardiovascular risk associated with obstructive sleep apnea. Studies have demonstrated that hypoxia reduces delta and REM sleep and prolongs stages 1 and 2 of NREM sleep (Berssenbrugge et al. 1984). However, it is not possible to determine whether the effects of obstructive sleep apnea in humans are caused by hypoxia per se or whether they are due to the frequency of arousals induced by breathing alterations during sleep or both (Berssenbrugge et al. 1983). To address this issue, we have been conducting studies using intermittent hypoxia in rats. As described by Perry et al. (2007), subchronic exposure to intermittent hypoxia (10% O 2 ) did not affect the biochemical blood parameters related to cardiovascular risk, but subchronic PSD reduced triglyceride. These parameters remained reduced when PSD was combined with intermittent hypoxia. Further investigation into the association between hypoxia and sleep deprivation, which is known as sequelae of obstructive sleep apnea, may be useful in understanding the influence of both factors on the cardiovascular system.

STRESS
In addition to sleep deprivation, achieved by either the single platform method or variations, the techniques induce high levels of stress and have been a subject of debate. If sleep is essential for health and life, then sleep deprivation is a biological stressor. Indeed, some signs of sleep deprivation are weight loss, reduced thymus weight (Coenen and Van Luijtelaar 1985) (Ledoux et al. 1996). Inc binding of CRF receptors in several brain areas duced hypothalamic CRF content is observed in deprived rats (Fadda and Fratta 1997). As the sleep cycle is distinctly affected by stress (Papale et al. further research is warranted to elucidate both how is interrupted by stress and the long-term effects in by sleep deprivation, an inherent stress condition, organism's coping skills. DEPRESSION Since its first experimental demonstration (Pflu Tölle 1971), the antidepressant effect of sleep vation for one night has been widely investigate ter many years, the procedure of sleep deprivati proven its efficacy for alleviating depression in a imately 60% of the cases after a single session, almost 90% after three sessions performed at one intervals (for review see Wu andBunney 1990, Justice andVan der Hoofdakker 1999). Howev effect is normally only transient, and, in most relapse occurs after the first episode of recovery (Southmayd et al. 1990). Despite the numerous of total sleep deprivation, and based on the theo PS pressure is increased in depressed patients (V al. 1980), some attempts were made to evaluate fects of selective PSD.
Apart from therapeutic PSD, the relationsh tween PSD effects and noradrenergic, serotonerg other effects of antidepressant drugs has been i gated. In 1983, Tufik and Silveira-Filho demon that PSD rats were more sensitive to imiprami fluoxetine and also to nomifensive, which showe phasic effect. As has been suggested, PSD may u the clinical effects of antidepressants in the con depression (Vogel 1975). Thus, potentiation of th apeutic effects of antidepressants by PSD sugge the depressants in clinical, as well as in locomo tivity (as used in this report) in rodents, share a mon mechanism (Tufik and Silveira-Filho 1983 development of anhedonia, characterized by loss The results indicate that PSD promotes a progressive decrease in daily sucrose intake, whereas the increase in sucrose intake observed during the rebound period could be attributed to a change in the strategy of coping. The majority of investigators have concluded that the immediate effect of a single session of sleep deprivation in humans is not influenced by the administration of antidepressant drugs (for review see Giedke andSchwärzler 2002, Adrien 2002). The common neurobiological mechanisms affected by pharmacological antidepressants and sleep deprivation suggest that sleep loss in some insomniac or depressed patients might be an endogenous compensatory process that is therapeutic rather than pathological. This proposal should open up new strategies for the treatment of depression (Adrien 2002).

CONCLUSION
Over the past few decades, a remarkable explosion of research has allowed us to construct a much more complete picture of the neurotransmitter, genetic, cellular, neurophysiological and behavioral changes that are affected by sleep deprivation.
Sleep deprivation seems to disrupt vital biological processes necessary for cognitive function and physical health, yet the ways in which the body is compromised are not fully understood. There is a pressure in modern society to carry out an increasing variety and number of activities during wakefulness. The expectation that these activities should be achieved tends to push sleep into background. This trends toward sleep deprivation and irregular sleep-wake patterns with resultant impaired concentration and memory reduces the quality of life and the ability to enjoy and complete activities. The balance may need to swing back toward awareness that adequate and regular sleep is required to promote a state of well-being during wakefulness. Once aware of this relationship, it becomes the responsibility of each individual to select his/her own combination of sleep and wakefulness by prioritizing the opportunities that present themselves every day.