|Home | About | Journals | Submit | Contact Us | Français|
Recent transcript profiling and microarray studies are beginning to unveil some of the mysteries of sleep. One of the most important clues has been the identification of the ER resident chaperone, BiP that increases with sleep deprivation in all species studied. BiP, an ER resident chaperone is the key cellular marker and master regulator of a signaling pathway called the ER stress response or unfolded protein response. The ER stress response occurs in 3 phases. It is healthy, protective and adaptive when the ER stress is moderate. Failure of the adaptive response leads to the activation of an inflammatory response. When the ER stress burden is great and prolonged, executioner pathways are activated. Collectively this work provides new evidence that modest sleep deprivation induces cellular stress that activates an adaptive response. Aging tilts the response to sleep deprivation from one that is adaptive and protective to one that is maladaptive. Understanding the pathways activated by sleep loss and the mechanisms by which they occur will allow the development of therapies to protect the brain during prolonged wakefulness and specifically in sleep disorders including those associated with aging.
Sleep is a ubiquitous phenomenon that most organisms experience yet our understanding of sleep regulation and function is at the early stages. Much is known about the circuitry involved in sleep and we are beginning to get glimpses into the cellular and molecular correlates of sleep largely through subtractive hybridization, differential transcript and cDNA microarray studies carried out in several species. Recent gene expression studies in rat1, mouse2, 3, Drosophila4, 5, and sparrow6 have all contributed to our growing understanding of the molecular changes that occur with sleep, wakefulness and extended wakefulness/sleep deprivation. Amongst the large number (~ 3,000) of genes that change with sleep state, change in one particular class of genes, the heat shock proteins/chaperones, is highly conserved in all species studied. One gene within this class called BiP has been shown to increase with sleep deprivation/extended wakefulness in all species examined - Drosophila7, rat cerebral cortex1, 8, 9, mouse cerebral cortex2, 3 and in the telencephalon of the white crowned sparrow6. BiP (Immunoglobulin Binding Protein) is also known as GRP78 (Glucose regulated protein 78) and under the most recent nomenclature is known as HSP5A (Heat shock protein 5A). This chaperone serves as a sentinel for up-regulation of several signaling pathways collectively called the unfolded protein response (UPR) or the endoplasmic reticulum (ER) stress response. The UPR responds to cellular stress and the response has 3 phases. It is healthy, protective and adaptive when the ER stress is moderate. When adaptation fails, genes that mediate host defense are activated. Persistent and prolonged stress activates executioner pathways.
This review will provide an overview of how the ER as a signaling organelle responds to stress through a set of co-coordinated pathways called the unfolded protein response. The initial cytoprotective response of the unfolded protein response and the later pro-apoptotic signaling pathway will also be discussed. The review will also summarize data from recent sleep deprivation and intermittent hypoxia studies that illustrate a role for the ER stress response in sleep. Finally, the relevance of these processes to neurodegenerative diseases is discussed. Understanding the pathways activated by sleep loss and the mechanisms by which they occur will allow the development of therapies to protect the brain during prolonged wakefulness and in sleep disorders including those associated with aging.
The endoplasmic reticulum is a sub-cellular organelle comprised of a reticular membranous network that extends throughout the cytoplasm and is contiguous with the nuclear envelope. It is the site where all secretory and integral membrane proteins are folded and post-translationally modified in ATP dependent chaperone mediated processes. The ER is also the site of steroid, cholesterol and lipid biosynthesis. It is the main storage site of calcium in the cell and is the major signal-transducing organelle in the cell that continuously responds to environmental cues to release calcium10.
In addition to being the most abundant chaperone in the ER, BiP is the first chaperone that a nascent peptide, destined for the ER, encounters during protein translation when it comes off the ribosome11. BiP binds to the hydrophobic domains on the peptide to prevent misfolding while the rest of the protein is being synthesized. Once the entire peptide is synthesized, it is folded with the aid of BiP and other chaperones; if properly folded it is escorted out of the ER. In contrast, if a protein is misfolded, it is refolded with the help of BiP, or if that is not possible, it escorted by BiP to the cytosol and proteasome for degradation12.
Because of its reticular nature the ER can sense and transmit signals that originate in any cellular sub-compartment. Perturbations that alter ER homeostasis disrupt folding of proteins and lead to the accumulation of unfolded proteins and protein aggregates, which are detrimental to cell survival. These perturbations include disturbances in calcium homeostasis, redox status, elevated secretory protein synthesis, and glucose or energy deprivation. As a consequence, the cell has evolved an adaptive coordinated response to limit accumulation of unfolded proteins in the ER and restore normal ER function. This signaling pathway is termed the ER stress response or the unfolded protein response (UPR) [for more extensive reviews see13–16]. The response to ER stress has 3 stages, adaptation, alarm and apoptosis, depending on the duration and intensity of the stress (see Figure 2).
On a cellular level, the UPR triggers three kinds of adaptive, protective cellular responses: I) up-regulation of ER chaperones such as BiP/GRP78 II) attenuation of protein translation and III) degradation of misfolded proteins by the proteasome by a process called ER associated degradation (ERAD) – (see schematic in Figure 1). These three responses are protective measures to reduce protein load and alleviate ER stress. When adaptation fails, ER initiated pathways signal alarm by activating the transcription factor, NF-KB that induces genes expressing mediators of host defense17. Lastly, excessive and prolonged stress leads to a maladaptive response and apoptosis18. When cell protective changes mediated by the UPR fail to restore folding capacity, CHOP (C/EBP homologous protein), c-jun NH2 terminal kinase (JNK) and caspases (discussed below) have been implicated in mediating apoptosis in response to sustained ER stress (see schematic in Figure 3).
BiP is a peptide-dependent ATPase and member of the heat shock 70 protein family that binds transiently to newly synthesized proteins translocated into the ER, and more permanently to underglycosylated, misfolded, or unassembled proteins. Under non-stressed conditions, BiP binds to the luminal domains of the three UPR transducers, IRE1 (Inositol Requiring 1), PERK [PKR (RNA-dependent protein kinase)-like ER kinase], and ATF6 (Activating Transcription Factor 6) to maintain them within the ER (see figure 1). Upon accumulation of unfolded proteins, BiP is sequestered from ATF6, IRE1 and PERK to chaperone the misfolded proteins thereby permitting the activation of one or more of these transducers19.
PERK is a type I transmembrane serine threonine kinase that appears to be present in most cells. It is held in an inactive monomeric state by binding to BiP. When this binding is disrupted, PERK homodimerizes and phosphorylates itself, thereby activating itself and initiating its eIF2α (eukaryotic Initiation Factor 2α) kinase activity. Phosphorylation of eIF2α causes a general decrease in translation of most proteins. However some selected proteins such as ATF4 and BiP, with internal ribosomal entry sites (IRES), are translated more efficiently20 and hence their protein levels actually increase. Other kinases including GCN2 (general control of nitrogen metabolism kinase 2), HRI (heme-regulated inhibitor kinase) and PKR also inhibit translation through eIF2α phosphorylation, however they do so by mechanisms other than ER stress21.
ATF6 is a 90-kDa bZIP protein that is activated by posttranslational modifications. ATF6 activation as part of the UPR leads to its translocation to the Golgi and cleavage by site-1 protease (S1P) and S2P. The 50-kDa cleaved ATF6α translocates to the cell nucleus, where it binds to the ER stress response element CCAAT(N)9CCACG22 in genes encoding ER chaperone proteins such as BiP and GRP94. GRp94 is a member of the heat shock90 family of chaperones. This results in increases in the level of these proteins and hence increasing protein folding activity in the ER22;23
Once activated, IRE1α gains endoribonuclease activity and splices X-box binding protein (XBP) 1 mRNA, generating a spliced variant (XBP1s) that functions as a potent transcriptional transactivator of genes involved in ER expansion, protein maturation, folding and export from the ER, as well as export and degradation of misfolded proteins24–28. IRE1α may also degrade ER-targeted mRNAs, thus providing another mechanism to decrease the production of new proteins in the organelle29, 30.
C/EBP homologous protein (CHOP) also known as growth arrest and DNA damage 153 (GADD 153), together with Bcl-2 (B-cell lymphoma protein 2) family members [Bak (BCl-2 homologous antagonist/killer)/Bax (BCl-2-associated X)], caspase-12 and c-jun NH2 terminal kinase (JNK) are components of the ER stress mediated apoptotic pathway31. CHOP is a transcription factor of the C/EBP family. Under non stress conditions the basal expression of CHOP is low. However CHOP levels have been shown to increase markedly when ER stress is not alleviated and persists32. Some of the mechanisms by which CHOP mediates apoptosis include I) the inhibition of protective anti-apoptotic factors like Bcl-2 and perturbation of cellular redox state by depletion of the anti-oxidant glutathione33; II) promotion of apoptotic caspase activity33 and III) the translocation of Bax from the cytosol to the mitochondria34.
Insertion of Bax into the mitochondrial membrane is essential for cytochrome c release and mitochondrial mediated apoptosis35. Bax translocation to the mitochondria is also downstream of JNK. The caspase family of cysteine proteases is a key mediator of programmed cell death or apoptosis. Murine caspase-12 which is a member of the interleukin-1β converting enzyme subfamily of caspases36 is an initiator caspase and the central player in ER induced apoptosis18. Once activated, caspase-12 translocates from the ER to the cytosol where it cleaves caspase-9, which induces the cleavage of caspase-3, the executioner caspase, in a cytochrome c-independent manner and activation of the rest of the apoptotic pathway37.
In addition to handling protein load, the adaptive response of the UPR serves both anti-apoptotic and anti-oxidant roles. The anti-apoptotic molecules in the ER identified thus far are the chaperones BiP, calreticulin, protein disulfide isomerase (PDI), and DAD1 (Defender Against apoptotic cell Death)38, 39. It has been shown that over expression of BiP leads to suppression of CHOP and CHOP mediated cell cycle arrest and apoptosis40–42 as well as inhibition of p38 MAPK mediated apoptosis43. Other studies have also suggested that a pool of BiP serves an anti-apoptotic role by forming complexes with the caspases (caspase-7 and caspase-12) at the ER surface preventing their activation and release44, 45. Also associated with the ER are the anti-apoptotic Bcl-2 family members Bcl-2 and Bcl-XL which inhibit cytochrome c release from the mitochondria during apoptosis.
PERK-dependent signaling, independent of translational inhibition, leads to the activation of the pro-survival transcription factor NRF-2 (NF-E2-related factor-2) via site-specific phosphorylation46. This PERK dependent phosphorylation leads to nuclear localization of NRF-2 and transcription of genes with the anti-oxidant response element. NRF-2 activation contributes to the maintenance of glutathione levels, which in turn functions as a buffer for the accumulation of reactive oxygen species during the unfolded protein response47. NRF-2 transcript has been shown to increase with sleep deprivation consistent with the idea that there is an increase in ROS with sleep deprivation and increased energy use2, 48.
Prolonged wakefulness over a period 6h or more does induce ER stress and the UPR in mouse cerebral cortex49. All of the components of the unfolded protein response pathway occur with sleep deprivation. Protein levels of BiP increase with 6, 9 and 12h of prolonged wakefulness. Immunoprecipitation studies indicate that BiP is dissociated from the kinase PERK with 6h of sleep deprivation. In addition the phosphorylation of PERK and eIF2α indicate inhibition of protein translation. Ribosome profiles from mice that have been subjected to 6h of sleep deprivation indicate that there is a disaggregation of polysomes into monosomes in the sleep deprived animals suggesting an attenuation of protein translation49. Active translation is associated with greater amounts of polysomes as was observed in undisturbed control animals.
The concept of the unfolded protein response had been suggested by others from microarray studies that showed increased cortical mRNA expression of the UPR associated chaperones BiP and GRP94 in rats during wakefulness1. In addition, PERK cerebellar transcript levels were also found to be higher in wakefulness than in sleep1. Other UPR specific transcripts that change with sleep deprivation include DNA-J which is a co-chaperone of BiP, XBP-1, calreticulin, caspase-9, ATF4 and ATF6 (see Table 1 for all genes that change with sleep deprivation).
Immunohistochemical studies also indicate that there is an increase in protein levels of ER chaperones GRP94, BiP/GRP78 and ERp72 during sleep deprivation in both the dorsal and lateral cortex of mouse brain9 again compatible with the induction of the unfolded protein response. It seems likely that the unfolded protein response with sleep deprivation might occur in other brain regions, and whether this is so will need to be addressed in future studies. Interestingly, it has recently been shown that BiP transcript levels increase in the liver of the AK mouse strain with sleep deprivation3. Further studies need to be carried out to ascertain whether sleep deprivation through the UPR affects peripheral organs. This study also demonstrated that BiP increased with sleep deprivation in the cortices of the three mice strains C57, AK and DBA studied while other UPR components, XBP-1, DNAJ and calreticulin increased only in the cortices of AK and DBA mice3.
The induction of the UPR by sleep deprivation also does not seem unique to mice since rest (sleep) deprivation in Drosophila, a state analogous to sleep in mammals7, 50, 51 also leads to up-regulation of the mRNA for BiP7, 51. Further, protein levels of BiP also increase two-fold, in Drosophila brain, with as little as 3h of sleep deprivation52. While it is known that PERK and XBP-1 are found in Drosophila it is not known whether these molecules change with sleep deprivation in a manner similar to those in mammals. Thus, further investigation of the UPR in Drosophila is needed.
This recent work on sleep deprivation/sleep on the unfolded protein response provides insights into earlier studies by Nakanishi et al53 and Ramm and Smith54, that demonstrated that protein synthesis was enhanced during sleep. These earlier studies lend credence to the findings that sleep deprivation is associated with attenuation in protein translation, with likely increased protein synthesis during recovery sleep. A recent preliminary study by Ding and colleagues using SELDI mass spectrometry also indicates that protein expression is decreased with sleep deprivation55. Thus, alteration in protein translation is a key consequence of sleep deprivation with recovery of this process during sleep.
The increase in BiP expression with sleep deprivation in Drosophila is followed by a decline/diminution in BiP with recovery sleep. In Drosophila, BiP protein levels return to baseline levels over 24h with recovery sleep following an almost 3-fold increase with 6h of sleep deprivation52. While this occurs, these data do not indicate whether this is simply a consequence of sleep deprivation or whether alteration in BiP levels affects sleep behavior.
To address this, BiP transgenics, one with higher levels of BiP than normal and one with inactive BiP, have been used to examine the effect of BiP levels on baseline sleep and recovery sleep52. Altering BiP expression levels did not change the amount of baseline sleep the flies acquired. Changing the amount of functional BiP did however alter the amount of recovery sleep in flies following sleep deprivation. Flies with high levels of BiP had more recovery sleep than flies with normal BiP levels, i.e., 30% more sleep in the first 4h post deprivation. In contrast, flies with inactive BiP had less recovery sleep than flies with normal sleep, i.e., 60% less sleep in the first 4h post deprivation. Thus, altering BiP impacts recovery sleep and BiP itself may be sleep promoting. However, high levels of BiP repress the UPR as all 3 transducers of the UPR remain bound to BiP when there is excess BiP56. These transducers need to become free for the UPR to occur. It may be that with the protective UPR pathway repressed, more sleep is required to recover from the ill effects of sleep deprivation. In contrast, inactive BiP leads to an earlier induction of the UPR. Thus, the need for recovery sleep may be reduced.
Whether the altered amounts of recovery sleep when BiP levels are manipulated is due to BiP itself or more indirectly through the other effects on the UPR just described remains to be determined.
In mammalian systems as well, there is evidence of an effect of these pathways on sleep/wake control. A study from the McGinty laboratory hypothesizes that sleep homeostasis may be induced by an increased demand for brain protein synthesis57. This hypothesis builds on the model that increasing protein synthesis may be a function of sleep and that some signals reflecting the protein synthetic load affect sleep regulation. As described earlier in the review, phosphorylation of eIF2α during ER stress leads to inhibition of protein translation. In their study Methippara et al58 used the small molecule inhibitor salubrinal that prevents dephosphorylation of eIF2α to test whether a build up of p-eIF2α had any effect on sleep58. Salubrinal administered ICV for 12h significantly increased SWS2 and suppressed active wake for the first 3h. For the remaining 9h, salubrinal did not affect sleep/wake. During the first 3h, salubrinal increased the number of SWS2 episodes and also increased delta power. The authors also reported that salubrinal treatment for 3h significantly increased the total number of p-eIF2α labeled neurons in the basal forebrain region. In addition, they found that administration of salubrinal activated GABAergic rostral median preoptic nucleus neurons as indicated by fos expression57. Taken together, this data suggests that ER stress signaling through phospho-eIF2α may play a role in sleep control and suggests that sleep is part of the compensatory response to ER stress in the brain.
The adaptive UPR response to sleep deprivation is attenuated in aged mice (22–24 months of age). All 3 components of the ER stress response/UPR, to sleep deprivation are compromised with aging. Specifically, ER chaperones like BiP are not increased with sleep deprivation in old animals, protein translation is not attenuated, and ER associated proteasomal degradation is reduced59. Following 3, 6, 9, and 12h of sleep deprivation BiP levels did not increase in the cerebral cortex of aged mice in contrast to the robust increase in young animals even at baseline (i.e., without sleep deprivation). BiP levels were 30% less in aged animals compared to young (3 month old) mice suggesting that at baseline protein folding is compromised in older animals. In older animals levels of activated PERK were low and there was also no significant activation of eIF2α by PERK with sleep deprivation. Thus, inhibition of protein translation that occurs during the UPR with sleep deprivation in young mice does not occur in the aged animals. Further, levels of GADD34 (Growth Arrest and DNA Damage-inducible protein) a protein phosphatase 1 (PP1)-interacting protein, which causes PP1 to dephosphorylate eIF2α, and relieve the translational block imposed by eIF2α phosphorylation18, 60 were increased both with age and sleep deprivation. Both this and the lack of translational inhibition suggest that protein translation proceeds in aged animals despite the ER stress and that there is most likely an increase in misfolded proteins and possibly in protein aggregation.
Finally, while baseline ubiquitin levels were higher in aged animals compared to young animals, the aged animals did not display further increases in ubiquitin with sleep deprivation, whereas they increased in the young. This suggests that misfolded proteins were not being cleared by degradation in the aged animals. In eukaryotic cells, the ubiquitin-proteasome system (UPS) is the main pathway for eliminating misfolded proteins61, 62. Misfolded or damaged proteins are tagged for UPS mediated degradation by the covalent attachment of ubiquitin molecules61.
Coupled with the absence of a robust protective UPR response, the pro-apoptotic signaling pathway alluded to earlier (see Figure 2) is turned on in older mice, even at baseline, and is much further increased by sleep deprivation (Figure 3). There was a dramatic increase in the pro-apoptotic factor CHOP with age that was further increased in older animals with prolonged waking. As described earlier, a role for CHOP in mediating apoptosis in response to ER stress is well established32, 63. Increased CHOP expression results in down-regulation of BCl-2 expression, depletion of cellular glutathione, and exaggerated production of reactive oxygen species33. Thus anti-oxidant systems are suppressed and free radical production is increased in the presence of CHOP.
The caspase cascade, that is a key mediator of programmed cell death, was also activated in aged animals. Levels of activated cleaved caspase-12 were higher in aged animals and were further increased with sleep derivation. Caspase-12 is thought to be key for ER mediated cell death. The up-regulation of CHOP, GADD34 and caspase-12 are all indicative of sustained ER stress. When ER stress is not alleviated by the adaptive arm of UPR, the pathway switches to a mechanism that promotes cell death.
Another group of molecules called the 14-3-3 proteins which play an anti-apoptotic role also change with sleep deprivation in both young and aged animals64. The 14-3-3 proteins prevent apoptosis through sequestration of Bax and other pro-apoptotic factors like Bad (BCl-2-antagonist of cell death) and ASK1 (Apoptosis signal-regulating kinase 1)65. During cellular stress Bax dissociates from 14-3-3 and relocates to the mitochondria. The decrease in 14-3-3 isoforms with 6h of sleep deprivation64 is indicative of cellular stress and an increase in pro-apoptotic signaling.
Aged animals exhibit more fragmented sleep59, 66, 67. Surprisingly, recovery sleep following sleep deprivation is less in older animals than young. This has been shown in humans68, 69 and in rats67, 70. Given that sleep deprivation in older animals is more injurious, one might have expected more, not less, recovery sleep. That older animals, who exhibit reduced recovery sleep following sleep deprivation, have no BiP response to sleep deprivation is compatible with the concept that BiP itself may play a direct role in determining the magnitude of sleep recovery, as discussed earlier.
While acute sleep deprivation leads to induction of the UPR, it appears that chronic sleep loss or long term sleep deprivation as assessed by an increase in BiP transcript levels does not. A study by Cirelli et al71 showed that rat cerebral cortex BiP mRNA levels do not increase after long-term (7 days) sleep deprivation as much as after short-term (8h) sleep deprivation. It is likely that 7 days of sleep deprivation results in sustained ER stress much like that experienced by aged animals which results in a shutdown of the adaptive ER stress response. It is likely that Thus, it is likely to be more injurious. Further studies are needed to address this.
A more recent study indicates that long term REM sleep loss (6–10 days) does lead to increased apoptosis in several regions of rat brain72. Using amino cupric staining as a marker of neuronal degeneration, TUNEL (TdT-mediated dUPT nick end labeling) assays and the ratio of Bcl-2 to Bax as indices of apoptosis this study demonstrates that neurons in locus coeruleus (LC), laterodorsal tegmentum (LDT), medial preoptic area (MPO) but not in the lateral septum undergo degeneration with REM sleep loss. The increase in Bax positive neurons over BCl-2 positive neurons observed in this study suggests that the last apoptotic phase of the UPR is activated in the REM sleep deprived animals. The absence of any apoptotic factors in the lateral septum following REM sleep loss illustrates nicely that there is differential vulnerability between neuronal groups to stress.
Producing exposure to intermittent cyclical hypoxia/reoxygenation similar to what occurs in obstructive sleep apnea, results in ER stress and injury in select brainstem motor neurons73. Many motor-neurons process large amounts of secretory and membrane proteins that must be properly folded within the ER74, and as such are prone to experiencing ER stress and the UPR. Short term intermittent hypoxia (3 days) selectively activates the PERK pathway of the UPR in some motor nuclei including hypoglossal and facial but not the motor trigeminal. Longer term intermittent hypoxia for 8 weeks results in similar PERK activation by phosphorylation in both the hypoglossal and facial motor neurons. In addition to PERK activation, however, the pro-apoptotic proteins, CHOP, GADD34, cleaved caspase-7 and caspase-3, are all increased with longer term intermittent hypoxia in these motor neurons. As a result these motor neurons undergo UPR even at baseline that is exacerbated by long-term intermittent hypoxia. The presence of CHOP and subsequently GADD34 lead to dephosphorylation of eIF2α, increased protein synthesis and greater ER stress in these neurons. The inability of these neurons to relieve the ER stress leads to neural injury that is observed at the ultrastructural level. Specifically, the ER is swollen and distorted and there is disaggregation of ribosomes and degranulation of rough ER73. In contrast, the trigeminal neurons in which eIF2α is phosphorylated and the adaptive response of the UPR is turned on have no such injury. Use of salubrinal, a small molecule inhibitor of eIF2α dephosphorylation, protected the susceptible motor neurons against neural injury. With this protein, translation is inhibited. Salubrinal treatment also decreased the amount of the pro-apoptotic factor CHOP. Thus, there is a differential response and ability to handle ER stress by different motor neurons. Neurons with low levels of baseline CHOP are able to mount a UPR that confers protection to intermittent hypoxia, while those motor neurons that experience sustained stress as evidenced by increased CHOP expression are not able to mount an adaptive UPR, and subsequently are more susceptible to ER stress and injury (See model in Figure 4). This effect of cyclical intermittent hypoxia may occur in other tissues and this is an area of future study.
Misfolded proteins and the associated ER stress are emerging as common features of neurodegenerative diseases (see reviews39, 75). Neuronal loss in both familial and sporadic forms of neurodegenerative disorders is often accompanied by aggregation of misfolded proteins76. Accumulating evidence suggests that ER dysfunction and aberrant protein degradation play a role in dopamine neuron loss in Parkinson's disease77. The ER stress response has also been implicated in Alzheimer's disease, amyotrophic lateral sclerosis, Huntington's disease and spinocerebrallar ataxias39, 75, 77. Accumulation of misfolded proteins that lead to alterations in organelle structure including the ER has been described in transgenic models of ALS, Alzheimer's and Huntington's disease39; 78. Neurons over expressing mutant presenilin 1 have been shown to be more sensitive to ER stress induced apoptosis79. Whether or not the mutation in presenilin 1 down-regulates BiP and induces CHOP is still a matter of debate. It has also been shown that extended polyglutamine tracts also stimulate ER stress induced cell death80, 81. These diseases usually appear late in life and are associated with aging. It is well documented that patients with neurodegenerative diseases like Parkinson's and Alzheimer's experience sleep fragmentation82, 83 84. Sleep fragmentation could place an additional burden on an already stressed protein folding and degradation system and could exacerbate protein aggregation acting as a positive feedback in these conditions. Thus, one needs to question whether sleep fragmentation in neurodegenerative disorders accelerates their progress based on the mechanism described in this review. This is, I believe, a topic worthy of further study.
Both sleep deprivation and the cyclical intermittent hypoxia that occurs in sleep apnea lead to ER stress. During ER stress the protein production machinery is compromised and protein misfolding occurs. The ER responds to this stress by up-regulating a series of coordinated cellular protective signaling pathways called the unfolded protein response. This protects the cell from the effects of misfolded proteins that can form toxic protein aggregates. This response appears to be highly conserved as sleep deprivation up-regulates BiP, the sentinel marker of the UPR in all species studied thus far, including fruit flies, birds and rodents. The protective response can be overwhelmed by additional stress and if the cumulative burden is too great then pro-apoptotic signaling is activated. This occurs in older animals with short-term sleep deprivation, in certain motor neurons with cyclical intermittent hypoxia and with prolonged sleep deprivation even in young animals. These are the beginning of pharmacological approaches to modify these pathways and this will be an area of further study.
I am grateful to Drs Allan Pack and Sigrid Veasey for their helpful comments during the writing of this manuscript. I would like to thank Ms. Jennifer Montoya for the illustrations and Mr. Daniel Barrett for editing. This study was supported by NIA grants AG025353 and AG17628.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.