Evidence suggests that IL-1β is involved in promoting physiological nonrapid eye movement (NREM) sleep. IL-1β has also been proposed to mediate NREM sleep enhancement induced by bacteria or their components. Mature and biologically active IL-1β is cleaved from an inactive precursor by a cysteinyl aspartate-specific protease (caspase)-1. This study aimed to test the hypothesis that inhibition in brain of the cleavage of biologically active IL-1β will reduce in rats both spontaneous NREM sleep and NREM sleep enhancement induced by the peripheral administration of components of the bacterial cell wall. To test this hypothesis, rats were intracerebroventricularly administered the caspase-1 inhibitor Ac-Tyr-Val-Ala-Asp chloromethyl ketone (YVAD; 3, 30, 300, and 1,500 ng) or were pretreated intracerebroventricularly with YVAD (300 ng) and then intraperitoneally injected with the gram-negative bacterial cell wall component LPS (250 μg/kg). Subsequent sleep-wake behavior was determined by standard polygraphic recordings. YVAD administration at the beginning of the light phase of the light-dark cycle significantly reduced time spontaneously spent in NREM sleep during the first 12 postinjection hours. YVAD pretreatment also completely prevented NREM sleep enhancement induced by peripheral LPS administration at the beginning of the dark phase. These results, in agreement with previous evidence, support the involvement of brain IL-1β in physiological promotion of NREM sleep and in mediating NREM sleep enhancement induced by peripheral immune challenge.
- interleukin-1 converting enzyme
- caspase inhibitors
although il-1β was originally described as a product of the peripheral immune system, there is now ample evidence that IL-1β, IL-1β receptors, and the IL-1β receptor antagonist are constitutively expressed in the normal brain (56). IL-1β modulates behaviors, such as feeding, sexual behavior, social exploration, locomotor activity, and sleep (46). IL-1β consistently has been shown in several animal species to enhance nonrapid eye movement (NREM) sleep and inhibit rapid eye movement (REM) sleep (38, 46). Moreover, central administration of the IL-1β receptor antagonist (41) or of antibodies directed against IL-1β reduces spontaneous NREM sleep in normal animals and inhibits the physiological NREM sleep rebound that follows sleep deprivation (42, 43). Mutant mice lacking the IL-1β type I receptor spend less time in NREM sleep than control mice (14). IL-1β mRNA expression in rat brain exhibits diurnal variation with levels higher during the light (rest/sleep) period than during the dark (active) period (9, 50). IL-1β mRNA increases during sleep deprivation (32). IL-1β is detected more frequently in plasma samples taken from humans during sleep than during waking (19). IL-1β-like activity in cerebrospinal fluid of cats varies in phase with the sleep-wake cycle (31), and IL-1β plasma levels in humans peak at sleep onset (36). IL-1β inhibits wake-active neurons in the hypothalamic preoptic area (POA) and the adjoining magnocellular basal forebrain (1), as well as serotonergic neurons in the dorsal raphe nucleus (33). IL-1β also affects different brain neurochemical systems involved in sleep regulation (22). Taken together, these data suggest that IL-1β is involved in physiological regulation and promotion of NREM sleep.
IL-1β has also been proposed to mediate changes in NREM sleep induced by bacteria or their components. Although the precise temporal responses to infection vary with pathogen and route of infection, acute infections are generally characterized by an increase in NREM sleep and an inhibition of REM sleep (46). Bacteria and their components induce IL-1β synthesis and release (46). Blockade of IL-1β binding to its cellular receptors (by means of the IL-1β receptor antagonist, an IL-1β-soluble receptor or an IL-1β receptor fragment) antagonizes NREM sleep enhancement induced by components of both gram-positive and gram-negative bacteria (23, 51).
IL-1β is synthesized in the cytoplasm as a 31-kDa prohormone that possesses little, if any, biological activity (11). The inactive prohormone is cleaved by cysteinyl aspartate-specific protease (caspase)-1 to the mature 17-kDa proinflammatory cytokine (3, 4, 10, 26). This protease, originally termed IL-1 converting enzyme (ICE; see Refs. 3–5, 26), is critically important in cytokine processing. Most caspases are involved in apoptosis, but human caspases-1, -4, and -5 and mouse caspases-11 and -12 seem to play an important role in inflammation and cytokine activation, with little contribution to cell death (48, 53). Caspases and their roles in numerous pathologies have been reviewed elsewhere (48, 49, 53, 55). Caspase inhibition reduces LPS- and Staphylococcus aureus-induced IL-1β release in septic patients (39). Caspase-1 inhibitors are now commercially available and have been shown to be effective in blocking or inhibiting IL-1β actions in a variety of models (e.g., see Refs. 25 and 47).
The aim of this study was to test the hypothesis that inhibition in brain of the cleavage of pro-IL-1β to biologically active IL-1β will reduce both spontaneous NREM sleep and NREM sleep enhancement induced by peripheral immune challenge. To test this hypothesis, rats were intracerebroventricularly (ICV) administered with the caspase-1 inhibitor Ac-Tyr-Val-Ala-Asp chloromethyl ketone (YVAD) or were pretreated ICV with YVAD and then intraperitoneally (IP) injected with gram-negative bacterial cell wall LPS. Subsequent sleep-wake behavior was determined. Results of the present study show that inhibition in brain of the cleavage of pro-IL-1β to biologically active IL-1β inhibits both spontaneous NREM sleep and NREM sleep enhancement induced by peripheral immune challenge.
MATERIALS AND METHODS
The caspase-1 inhibitor YVAD was purchased from Calbiochem-Merck (Darmstadt, Germany), dissolved in isotonic pyrogen-free saline (PFS; Abbot Laboratories, North Chicago, IL) containing 0.6% DMSO (Sigma, St. Louis, MO) and ICV administered in a volume of 1 μl for the three lower doses tested in the study (3, 30, and 300 ng) and in a volume of 2 μl for the highest dose tested (1,500 ng). YVAD is preferentially active on caspase-1 (Ki of 0.8 nM) compared with caspase-4 and -5 (Ki of 362 and 163 nM, respectively) (16). YVAD is an irreversible caspase-1 inhibitor. Caspase-1-like activity is reduced in cortical homogenates 24 h after YVAD administration and then returns to control levels 6 days later (47). LPS (Escherichia coli serotype O111:B4) was purchased from Sigma, dissolved in PFS, and injected in a volume of 2 ml/kg ip.
The experiments were performed on 34 male Sprague-Dawley rats (225–250 g at time of surgery; Charles River). The animals were anesthetized (pentobarbital sodium 40 mg/kg with chloral hydrate 180 mg/kg ip), injected with an analgesic (butorphanol tartrate) and a broad-spectrum antibiotic (penicillin G benzathine), and positioned in a stereotaxic apparatus. Stainless steel screws placed over frontal, parietal, and occipital cortices served as EEG and ground electrodes. A calibrated 30-kΩ thermistor (Omega Engineering, Stamford, CT) was implanted between the dura mater and the skull over the parietal cortex to measure brain cortical temperature (Tcort). Teflon-coated silver wires were inserted in the neck muscles to record electromyographic (EMG) activity. A polyethylene cannula was stereotaxically implanted into the lateral ventricle for ICV injections. Insulated leads were routed from the electrodes and the thermistor to a Teflon pedestal (Plastics One, Roanoke, VA) that was cemented in place with dental acrylic (Isocryl; Lang Dental Supply, Wheeling, IL) and used to connect the animal to the recording apparatus via a flexible tether and slip ring (Plastics One). At the end of surgery, the incision was treated topically with polysporin (polymixin B sulfate-bacitracin zinc), and the animals were placed under heat lamps and monitored until recovery from anesthesia.
The animals were allowed 7 days of recovery from the surgical procedures. On the third postsurgical day, the rats were connected to the recording apparatus (see Apparatus and Recording) for adaptation. Rats were housed individually in environmentally-controlled chambers maintained at 21 ± 0.5°C with a 12:12-h light-dark cycle. A standard rat maintenance diet (Mucedola, Settimo Milanese MI, Italy) and water were available ad libitum. Four days after surgery and at the end of the study, the location and free drainage of the ventricular cannulae were verified in vivo by means of the drinking response elicited by ANG II (200 ng icv) (13). Only animals that showed a positive response to ANG II at both times were included in this study. Two animals out of 36 did not respond to the second ANG II administration. Their results were discarded from the study. All procedures performed in these studies conformed to European Union (EEC Council Directive 86/609, OJ L 358,1; 12 December 1987) and Italian (D.L. n.116, G.U. Suppl. 40, 18 February 1992) laws and policies and were in accordance with the United States Department of Agriculture Animal Welfare Act and the National Institute of Health (Bethesda, MD) Public Health Policy on Humane Care and Use of Laboratory Animals.
Apparatus and Recording
Signals from the electrodes, as well as from the thermistor, were fed into a polygraph (model 7; Grass, Quincy, MA). The EEG was amplified (factor of 3,000) and analog bandpass filtered between 0.1 and 40 Hz (frequency response: ±3 dB; filter frequency roll off: 12 dB/octave). The conditioned signals were digitized with 12-bit precision at a sampling rate of 128 Hz (AT-MIO-64F5; National Instruments, Austin, TX). The digitized signals were stored as binary computer files until subsequent analyses.
Postacquisition determination of vigilance state was done by visual scoring of 12-s epochs using custom software. The animal’s behavior was classified as either wakefulness (W), NREM sleep, or REM sleep based on criteria previously reported (40). EEG power density values were obtained by means of fast-Fourier transform for each artifact-free 12-s scoring epoch for the frequency range of 0.5–20 Hz. Values in the 0.5-to-4.0 Hz (delta) frequency range were collapsed and integrated for 12-s epochs and used as measures of slow-wave activity (SWA) during NREM sleep. A minimum of 20 NREM sleep epochs per hour per rat was used to calculate hourly averages for SWA during NREM sleep (NREM SWA). If these criteria for a minimum number of NREM sleep epochs per hour were not met, SWA values for that animal per hour were not included in subsequent analyses.
Experiment 1: Effects of inhibition of caspase-1 on spontaneous sleep-wake behavior.
This experiment aimed to test the hypothesis that inhibition of caspase-1 in rat brain will reduce NREM sleep. For this experiment, rats (n = 19) were subdivided into two groups to assess the effects of YVAD administration at the beginning of both the light and dark phases of the light-dark cycle. Eleven rats were injected 15 min before light onset with PFS, PFS containing 0.6% DMSO [as vehicle (VEH) of the test substance] and four doses of YVAD (3, 30, 300, and 1,500 ng). No rat received more than three doses, and six to eight rats were injected for each dose. The effects of YVAD administration at light onset were investigated first because the hypothesis tested in the study was that YVAD reduces NREM sleep. Since rats spend more time in NREM sleep during the light than the dark phase of the light-dark cycle, decreases in NREM sleep should be more apparent during the light period. Eight animals were injected 15 min before dark onset with PFS, VEH, and, on the basis of the results obtained in animals injected at light onset (see above and results below), with two doses of YVAD (3 and 300 ng).
Experiment 2: Effects of inhibition of caspase-1 on LPS-induced changes in sleep-wake behavior.
This experiment aimed to test the hypothesis that inhibition of caspase-1 in the brain inhibits NREM sleep enhancement induced by peripheral LPS administration. Rats (n = 15) were subjected to a double-injection protocol. This protocol required four conditions. Condition 1: VEH ICV + VEH IP; condition 2: YVAD (300 ng) ICV + VEH IP; condition 3: VEH ICV + LPS (250 μg/kg) IP; condition 4: YVAD (300 ng) ICV + LPS (250 μg/kg) IP. All ICV injections (VEH and YVAD) were initiated 45 min before dark onset. The IP injections (VEH and LPS) were given 30 min later (15 min before dark onset).
All animals used in the study received the same volume of both vehicle and test substance, therefore serving as their own control. All administrations were randomly scheduled and made at least 4 days apart. Polygraphic recordings were begun at either light or dark onset (depending on the protocol) and were continued for 23 h.
For experiment 1, one-way ANOVA was used to determine whether values obtained after YVAD administration deviated statistically from values obtained after VEH injections. The duration of each vigilance state (NREM, REM, W), Tcort values, and values for SWA during NREM sleep were the dependent variables. The ANOVA analyses were conducted across 12-h time blocks, corresponding to the dark and light phases of the light-dark cycle. In each analysis, manipulation (VEH vs. YVAD doses) was the main (fixed) effect. If statistically significant differences were detected between manipulations, post hoc multiple comparisons were made using Fisher’s (protected) least significant difference test. For experiment 2, two-way ANOVA and Fisher’s post hoc test (when appropriate) were used to evaluate the interaction between treatments A (VEH or YVAD) and B (VEH or LPS). Analyses were conducted in 12-h time blocks for duration of each vigilance state (NREM, REM, W), Tcort values, and values for SWA during NREM sleep.
Administration of PFS containing 0.6% DMSO at both light and dark onset of the light-dark cycle did not significantly modify either sleep-wake behavior or Tcort compared with values obtained following PFS administration alone (data not shown). Because PFS containing 0.6% DMSO is the VEH for YVAD, all appropriate statistical comparisons were made to values obtained after this vehicle.
Experiment 1: Effects of inhibition of caspase-1 on spontaneous sleep-wake behavior
YVAD administration at light onset.
ANOVA analyses revealed significant effects of YVAD administration on sleep-wake behavior (Fig. 1, left, and Table 1). The amount of NREM sleep during the first 12 postinjection hours was reduced compared with values observed following VEH administration [F(4,451) = 4.76, P = 0.001]. Post hoc analyses revealed significant effects of the 3-, 300-, and 1,500-ng doses, with 30 ng not contributing to the overall effect. No dose-dependent effect was observed. No changes in NREM sleep compared with control condition (VEH administration) were observed during postinjection hours 13–23 following any of the YVAD doses administered. Changes in NREM sleep were mirrored in the first 12 postinjection hours by significant increases in wakefulness (Fig. 1, left and Table 1); F(4,451)=4.53, P = 0.001], with significant departures from VEH revealed for the 3-, 300-, and 1,500-ng doses. No significant changes compared with VEH administration were subsequently observed in wakefulness in postinjection hours 13–23. REM sleep and Tcort were not affected by any of the YVAD doses administered during postinjection hours 1–12 or 13–23. NREM SWA (Fig. 1, left and Table 1) was significantly reduced compared with control conditions by all of the YVAD doses tested in both the first 12 postinjection hours [F(4,447) = 15.1, P < 0.001] and during the 13–23 postinjection hours [F(4,400) = 30.8, P < 0.001]. No dose-dependent effects were revealed.
YVAD administration at dark onset.
Because no dose-related effects were observed when YVAD was administered at light onset, only two doses of YVAD (3 and 300 ng) were administered at dark onset (Fig. 1, right and Table 1). ANOVA revealed a significant effect of YVAD administration on the amount of time spent in NREM sleep [F(2,285) = 4.1, P < 0.02]. Post hoc analyses indicated this overall effect was due to a significant difference in the first 12 postinjection hours between changes induced by YVAD 300 and 3 ng and not to changes relative to vehicle administration. NREM sleep amount in the first 12 postinjection hours after administration of YVAD, 300 ng was significantly reduced relative to time spent in NREM sleep after administration of YVAD 3 ng. Overall, ANOVA for wakefulness indicated a significant effect of YVAD [F(2,285) = 4.5, P = 0.012], which was attributed to an increase in wakefulness after YVAD 300. Post hoc analyses also revealed differences between YVAD 300 ng and YVAD 3 ng in wakefulness. No differences in wakefulness amount were observed between the different experimental conditions in postinjection hours 13–23. REM sleep was inhibited by YVAD [F(2,285) = 3.9, P = 0.02], an effect post hoc analyses indicated was due to reduction in REM sleep after the YVAD 300-ng dose. Post hoc analyses also indicated that the amount of time spent in REM sleep differed between the two YVAD doses. No significant changes were observed in REM sleep in the subsequent 11-h time block. Tcort and NREM SWA were not affected by any of the YVAD doses during any time block that was analyzed.
Experiment 2: Effects of inhibition of caspase-1 on LPS-induced changes in sleep-wake behavior
In this experiment, animals were treated according to the double-injection protocol described in materials and methods. For the sake of brevity, three of the four experimental conditions are here identified as follows: VEH ICV + VEH IP = control condition, YVAD 300 ng ICV + VEH IP = YVAD alone, and VEH ICV + LPS 250 μg/kg IP = LPS alone.
First 12 postinjection hours.
(Fig. 2, left, shaded and Table 2). NREM sleep amount was unaffected by YVAD alone. IP LPS increased NREM sleep [F(1,416) = 19.2, P < 0.001], a response completely abolished when animals were pretreated with YVAD before LPS administration. These changes in NREM sleep were mirrored by changes in wakefulness, the amount of which was unaffected compared with the control condition by the administration of YVAD alone and was significantly reduced by administration of LPS alone. Pretreatment with YVAD completely antagonized changes induced in wakefulness by LPS. REM sleep amount, not significantly modified by YVAD alone compared with control condition, was significantly inhibited by LPS administration [F(1,416) = 47.9, P < 0001]. This reduction was potentiated by pretreatment with YVAD.
YVAD alone significantly reduced NREM SWA. LPS alone increased NREM SWA relative to YVAD alone. NREM SWA after the combination of YVAD and LPS was reduced relative to values obtained after LPS alone. However, values for NREM SWA were lower after each of the three treatment conditions than after the VEH conditions. Tcort was not altered by any of the treatments tested.
Second 12 postinjection hours.
NREM sleep amount was significantly reduced compared with the control condition by administration of both YVAD alone [F(1,394) = 27.2, P < 0.0001] and LPS alone [F(1,394) = 25.2, P < 0.0001], as well as by the combination of the two treatments [F(1,394) = 11.5, P = 0.0008]. Wakefulness was correspondingly increased in the same conditions. Compared with control condition, NREM SWA underwent the same changes described in the first 12 postinjection hours; it was significantly reduced by the administration of YVAD, LPS, and the combination of the two treatments, with values observed after LPS administration being significantly higher than those observed after the other two experimental manipulations (Fig. 2, right and Table 2).
Tcort was not affected by administration of YVAD alone, whereas it was increased by the administration of LPS alone. The LPS effect on Tcort was potentiated by pretreatment with YVAD.
The major findings of the present study are that inhibition of caspase-1 in the brain of rats inhibits spontaneous NREM sleep and NREM sleep enhancement induced by peripheral administration of LPS. These results, in agreement with previous evidence from other experimental models, directly support the involvement of IL-1β in physiological promotion of NREM sleep and in NREM sleep enhancement induced by immune challenge.
Experiments demonstrate that the administration of the IL-1β receptor antagonist (41) or of antibodies directed against IL-1β reduces spontaneous NREM sleep in normal animals (42). In addition, NREM sleep enhancement induced by muramyl dipeptide (a synthetic analog of muramyl peptides, the monomeric building blocks of bacterial cell wall peptidoglycan) is antagonized by the IL-1β soluble receptor (23) and an IL-1β receptor fragment (51). Each of these previously mentioned experiments interfered with the binding of IL-1β to its receptor. In the present study, a different approach was chosen, i.e., to target IL-1β upstream, inhibiting caspase-1 and thus preventing the cleavage of pro-IL-1β to its mature and biologically active form.
The effects of caspase-1 inhibition on spontaneous NREM sleep depend on the timing of administration. NREM sleep is inhibited when YVAD is administered at the beginning of the light phase, but not when administered at dark onset (Fig. 1). This interaction between the effects of caspase-1 inhibition and the light-dark cycle was expected, since IL-1β mRNA levels in brain undergo circadian fluctuations, being higher during the light than the dark phase (9, 50). For rats, the light phase also represents the rest phase, when they spend most time asleep. These observations suggest that it should be easier to detect the functional consequences of the inhibition of IL-1β activation and any possible NREM sleep reduction when IL-1β levels and NREM sleep amounts are physiologically higher.
The NREM sleep loss induced by YVAD administration at the beginning of the light phase of the light-dark cycle amounts to about 1 hr in the first 12 postinjection hours. This reduction in NREM sleep is quantitatively similar to that induced in rats by the ICV administration of antibodies directed against IL-1β (42). NREM sleep loss induced by caspase-1 inhibition seems not to require compensation, because it is not followed in the subsequent dark phase by any rebound increase. YVAD administration at the beginning of the dark phase does not affect the amount of spontaneous NREM sleep in postinjection hours 13–23, i.e., during the subsequent light phase (Fig. 1).
NREM sleep enhancement in response to peripheral LPS administration at the beginning of the dark phase of the light-dark cycle is a consistent finding of several studies in rats (24, 35, 45), as well as in other species (46). Peripheral LPS administration induces IL-1β mRNA expression and protein in the brain (2, 6, 17, 20). Results of the present study show that inhibiting caspase-1 from cleaving pro-IL-1β to form active IL-1β (i.e., upstream antagonism) blocks NREM sleep enhancement induced by LPS. These results are in agreement with, and complement previous observations showing that NREM sleep enhancement induced by components of the bacterial cell wall can be antagonized by preventing the binding of IL-1β to its cellular receptor [i.e., downstream antagonism (23, 51)]. Collectively, these data demonstrate that changes induced in NREM sleep by bacterial components are mediated, in part, by IL-1. These data also demonstrate that the role of IL-1β in mediating LPS-induced increases in NREM sleep is not restricted to occupancy of the IL-1β type I receptor by its ligand or by posttranscriptional events, but is also dependent on posttranslational processing.
Sleep, as any other behavior, is ultimately regulated by the central nervous system. Changes observed in sleep during bacterial infections or after the administration of components of the bacterial cell wall result from interactions between the central nervous and immune systems. This conclusion is directly supported by the results of the present study, which show that changes induced in rat NREM sleep by peripheral LPS administration are antagonized by inhibition of IL-1β in brain. Previous studies, in which ligands or receptors of both IL-1β (51) and TNF-α (52) were antagonized, also demonstrate that NREM sleep enhancement induced by immune challenge with bacterial products are blocked. As such, the conclusion that changes induced in sleep by immune challenge are mediated by the central nervous system is consistently supported by evidence obtained from different animal species, using different challenges (administered by different routes), targeting different cytokines and using different strategies for antagonism of cytokines.
SWA During NREM Sleep
When given at the beginning of the light phase, all the YVAD doses tested induce significant reductions in SWA during NREM sleep (Fig. 1). These reductions are long lasting, since they are evident during the entire 23-h recording period. As such, the duration of reduced SWA during NREM sleep exceeds the duration of reduced NREM sleep time, indicating a clear dissociation of these two process. The dissociation of responses to YVAD with respect to SWA during NREM sleep and the amount of time spent in NREM sleep demonstrates that different mechanisms are involved in the regulation of these processes. Data supporting differential regulation of SWA during NREM sleep and NREM sleep duration have also been derived from studies in which local application of IL-1β to the surface of the cerebral cortex increases SWA during NREM sleep without altering the amount of time spent in NREM sleep (57). Previous observations indicate that central administration at light onset of antibodies directed against IL-1β also inhibits NREM SWA (42). In addition, ICV administration of IL-1β increases NREM SWA (29, 41, 44, 54). It is now generally accepted that NREM SWA reflects sleep intensity (7, 8). Collectively, these data suggest that IL-1β impacts the intensity of NREM sleep, in addition to regulation of physiological NREM sleep duration. In contrast to responses to IL-1, the impact of LPS on NREM SWA is reportedly variable. Although LPS administration at light onset increases NREM SWA of rats (45), other studies demonstrate little or no effect of dark onset administration of LPS on this parameter (24, 30, 35, 45). In this present study, LPS administered at dark onset reduced NREM SWA. Although in this study both YVAD and LPS decrease NREM SWA, their effects are not synergistic or additive when administered together (Fig. 2). Additional studies are necessary to more fully elucidate the mechanisms mediating and functional consequences of the impact of caspase-1 inhibition and immune challenge on measures of sleep intensity.
Effects of caspase-1 inhibition on spontaneous REM sleep are different from effects on NREM sleep. YVAD administration at the beginning of the light phase does not alter REM sleep. REM sleep is inhibited when the highest dose of YVAD used in this study is administered at dark onset (Fig. 1, right). In healthy animals and humans, REM sleep is entered from NREM sleep. The amount of time rats spend in NREM sleep is less during the dark period. Under these conditions, even a modest additional reduction in NREM sleep could result in an REM sleep loss secondary to NREM sleep reduction. REM sleep is also inhibited by LPS injected at dark onset (Fig. 2), an observation in agreement with abundant evidence showing that infectious agents and their components, such as LPS (35), inhibit this sleep phase in rats and other species (46). The LPS-induced reduction in REM sleep is potentiated by pretreatment with YVAD. The mechanisms responsible for REM sleep reduction under conditions of caspase-1 inhibition and LPS challenge remain to be elucidated. Taken together, data in this study demonstrating different effects of capsase-1 inhibition on NREM and REM sleep suggest that regulation of the two phases of sleep are differently modulated by IL-1.
Inhibition of caspase-1 does not alter Tcort. This observation is in agreement with previous reports demonstrating that antagonism of IL-1β does not affect Tcort (41, 42). Therefore, in spite of its well-known role in inducing fever (12), IL-1β may not be involved in physiological thermoregulatory mechanisms. Under the conditions of this study, LPS administration at dark onset does not significantly alter Tcort during the first 12 postinjection hours. Depending on the dose and the serotype, different effects of LPS administration at dark onset on Tcort of rats have been described. LPS may have little (45) or no effect (30, 35) on Tcort, or there may be an initial decrease, followed by a long-lasting increase (24) in Tcort.
Caspase-1 Inhibition and its Effects
Beyond its role in posttranslational processing of IL-1, caspase-1 has been shown to cleave active IL-18 from its inactive precursor (34). As a secondary effect, since IL-1β stimulates the production of TNF-α and IL-6 (15, 18, 28), inhibition of caspase-1 may also impair production of these cytokines. Each of these cytokines has been implicated as a regulator and/or modulator of NREM sleep (46). This raises the issue of the possible involvement of other cytokines in mediating the effects of caspase-1 inhibition of IL-1β production on sleep-wake behavior.
IL-18 administration promotes NREM sleep in both rabbits and rats (27). The observation that ICV administration of antibodies directed against IL-18 does not affect spontaneous sleep-wake behavior of rabbits (27) suggests that any reduction of IL-18 levels induced by caspase-1 inhibition is not likely a mediator of changes in spontaneous sleep reported in this study. On the other hand, because antibodies directed against IL-18 attenuate NREM sleep enhancement induced by muramyl dipeptide (27), it is possible that a reduction in IL-18 by caspase-1 inhibition may play a partial role in mediating the antagonism of LPS effects described in this study. Data demonstrate (46) that TNF-α plays a role in both physiological NREM sleep regulation and in mediating NREM sleep enhancement induced by immune challenge. Because IL-1β stimulates TNF-α production, it is possible that a reduction in TNF-α levels secondary to a caspase-1-induced reduction in IL-1β contributes to sleep changes described in the present study. Evidence suggests that IL-6 may modulate sleep-wake behavior under conditions when this cytokine is elevated but may not play a role in the regulation/modulation of sleep-wake behavior of healthy animals because antagonizing IL-6 does not alter sleep (21). IL-6 knockout mice do respond to LPS administration differently than do genetically intact mice (37), raising the possibility that reductions in IL-6 secondary to caspase-1-induced reductions in IL-1β may contribute, in part, to the responses to LPS reported in this study.
In conclusion, this study demonstrates that interfering with enzymatic cleavage of pro-IL-1β to its mature and biologically active form reduces spontaneous NREM sleep of rats. These data provide additional evidence that IL-1β is involved in the regulation of NREM sleep. Data indicating that caspase-1 inhibition reduces spontaneous NREM sleep of rats only when administered at light onset corroborates previous studies demonstrating a diurnal variation in IL-1β mRNA in rat brain (9, 50); caspase-1 inhibition effectively reduces NREM sleep only when IL-1β levels in rat brain are highest. That caspase-1 inhibition in rat brain blocks increases in NREM sleep induced by peripheral administration of LPS supports the hypothesis that cytokines in brain mediate the alterations in sleep-wake behavior that follow immune challenge. Although additional studies are necessary to more fully elucidate central mechanisms mediating responses to peripheral challenge, this present study clearly indicates a role for IL-1β in these responses.
This work was supported, in part, by National Institutes of Health Grant MH-64843 (to M. R. Opp and L. Imeri), by the Ministero dell’Istruzione, dell’Università e della Ricerca Scientifica (to L. Imeri), and by the Department of Anesthesiology, University of Michigan (to M. R. Opp).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2006 the American Physiological Society