Homeostatic sleep regulation in the absence of the circadian sleep-regulating component: effect of short light-dark cycles on sleep-wake stages and slow waves

Abstract

Background: Aside from the homeostatic and circadian components, light has itself an important, direct as well as indirect role in sleep regulation. Light exerts indirect sleep effect by modulating the circadian rhythms. Exposure to short light-dark cycle (LD 1:1, 1:1 h light - dark) eliminates the circadian sleep regulatory component but direct sleep effect of light could prevail. The aim of the present study was to examine the interaction between the light and the homeostatic influences regarding sleep regulation in a rat model.

Methods: Spontaneous sleep-wake and homeostatic sleep regulation by sleep deprivation (SD) and analysis of slow waves (SW) were examined in Wistar rats exposed to LD1:1 condition using LD12:12 regime as control.

Results: Slow wave sleep (SWS) and REM sleep were both enhanced, while wakefulness (W) was attenuated in LD1:1. SWS recovery after 6-h total SD was more intense in LD1:1 compared to LD12:12 and SWS compensation was augmented in the bright hours. Delta power increment during recovery was caused by the increase of SW number in both cases. More SW was seen during baseline in the second half of the day in LD1:1 and after SD compared to the LD12:12. Increase of SW number was greater in the bright hours compared to the dark ones after SD in LD1:1. Lights ON evoked immediate increase in W and decrease in both SWS and REM sleep during baseline LD1:1 condition, while these changes ceased after SD. Moreover, the initial decrease seen in SWS after lights ON, turned to an increase in the next 6-min bin and this increase was stronger after SD. These alterations were caused by the change of the epoch number in W, but not in case of SWS or REM sleep. Lights OFF did not alter sleep-wake times immediately, except W, which was increased by lights OFF after SD.

Conclusions: Present results show the complex interaction between light and homeostatic sleep regulation in the absence of the circadian component and indicate the decoupling of SW from the homeostatic sleep drive in LD1:1 lighting condition.

Keywords: Circadian sleep regulation; Homeostatic sleep regulation; Short light-dark cycles; Sleep effect of light; Slow waves.

Fig. 1

Sleep–wake parameters during baseline LD12:12 (n = 6) and baseline LD1:1 (n = 5) conditions. Line A: wakefulness, line B: slow wave sleep, line C: REM sleep. a/1, b/1, c/1: hourly averages of the times spent in the given vigilance stage. Thin black line: LD12:12; thick grey line: LD1:1. a/2, b/2, c/2: hourly averages of the times spent in the given vigilance stage calculated from the 24-h data in LD12:12- and in LD1:1 conditions. a/3, b/3, c/3: summarized time spent in the given vigilance stage during the first- and the second half of the day (CT1-CT12 hours and CT13-CT24 hours) in LD12:12- and in LD1:1 conditions. a/4, b/4, c/4: summarized time spent in the given vigilance stage during the whole day (24-h) in LD12:12- and LD1:1 conditions. a/5, b/5, c/5: hourly averages of the times spent in the given vigilance stage averaged separately for the light- (CT1, CT3, etc.) and for the dark hours (CT2, CT4, etc.) in LD1:1 condition. Black bars: average of the whole period (bright hours vs. dark hours); grey bars: average in the separate hours. SEM values are not depicted on panel a/5, b/5 and c/5. Black and white bars at the x axis of the panels a/1, b/1 and c/1 show the illumination pattern in the different hours. Grey asterisks (*) indicate significant deviation from the corresponding control (LD12:12) data (a/1, b/1, c/1) or indicate significance difference between the data derived from different time periods (LD12:12 baseline vs. LD1:1; a/2-a/4, b/2-b/4, c/2-c/4) or showing significant difference between the values belonging to light vs. dark hours (a/5, b/5, c/5). Significance was tested with two-way ANOVA with time and treatment as factors, followed by Sidak's multiple comparisons test. Sleep–wake times in light vs. dark hours in the LD1:1 condition depicted on a/5, b/5 and c/5 were compared using Welch’s t-test. Significance levels: *—p < 0.05; **—p < 0.01; ***—p < 0.001. Data are expressed as mean ± S.E.M

Fig. 2

Effect of 6-h gentle handling total sleep deprivation on slow wave sleep and delta power during LD12:12 (n = 6) and LD1:1 (n = 5) conditions. The figure depicts only the data of the recovery period after SD (R1–R16 hours). a–e: LD12:12 data, panel F-J: LD1:1 data. a, f: Hourly average of slow wave sleep times during baseline (grey line) and after sleep deprivation (black line). b and g: Changes of normalized delta (1–4 Hz) power expressed as percentage of corresponding baseline. Grey lines: baseline; black lines: after SD. c and h: Slow wave sleep times (min/h) in the consecutive hours of the R1–R6 period. In LD12:12, the R1–R6 period consisted bright hours only (c) while in LD1:1, R1–R3-R5 hours were bright and R2-R4-R6 hours were dark (h). d and i Slow wave sleep times in blocks containing the sum of the times of three different hours taken from the baseline (hours 7, 9, 11 and hours 8, 10, 12) and from the recovery period (R1, R3, R5 and R2, R4, R6 hours). Panel e and j: summarized amount of slow wave sleep at the end of the light phase (hour CT12) during the baseline LD12:12 and LD1:1 conditions and after total SD (hour R6). The column pair at the right side show summarized slow wave sleep time at the end of the day during baseline and after total SD. Black and white bars at the x axis show the illumination pattern in the different hours. Significance was tested with two-way ANOVA with time and treatment as factors, followed by Sidak’s multiple comparisons test. Summarized SWS sleep times during the first- and in the second half of the day depicted on E and J were compared using Welch’s t-test. Significance levels: *, # - p < 0.05; **, ##- p < 0.01; ***, ### - p < 0.001. Data are expressed as mean ± S.E.M

Fig. 3

Effect of 6-h gentle handling total sleep deprivation on REM sleep during LD12:12 (n = 6) and LD1:1 (n = 5) conditions. The figure depicts only the data of the recovery period after SD (R1–R16 hours). Panel a–b: LD12:12 data, panel c–d: LD1:1 data. Panel a and c: hourly average of REM sleep times during baseline (grey line) and after sleep deprivation (black line). Panel b and d: summarized amount of REM sleep at the end of the light phase (hour CT12) during the baseline LD12:12 and LD1:1 conditions and after total SD (hour R6). The column pair at the right side show summarized REM sleep time at the end of the day during baseline and after total SD. Black and white bars at the x axis show the illumination pattern in the different hours. Significance was tested with two-way ANOVA with time and treatment as factors, followed by Sidak’s multiple comparisons test. Summarized REM sleep times during the first- and in the second half of the day depicted on C and D were compared using Welch’s t-test. Significance levels: *, # - p < 0.05; **, ##- p < 0.01; ***, ### - p < 0.001. Data are expressed as mean ± S.E.M

Fig. 4

Effects of light onsets regarding S-W times (a-c) and epoch numbers (d-f) in LD1:1 (n = 5) condition. S-W data were averaged in consecutive 6-min long bins using the data of all light switches during the 24-h recordings. Two bins were analyzed in the pre-switch period (dark hour 48–54 min and 54–60 min) and two bins for the post-switch period (bright hour 0–6 min and 6–12 min). Panel a and d: wakefulness; panel b and e: slow wave sleep; panel c and f: REM sleep. Significance was tested with two-way ANOVA with time and treatment as factors, followed by Sidak’s multiple comparisons test. Differences were checked between the different 6-min time bins before and after the light switch during baseline and after SD and between the baseline vs. after SD values regarding the same time period. Significance levels: * - p < 0.05; ** - p < 0.01; *** - p < 0.001. Data are expressed as mean ± S.E.M

Fig. 5

Effects of dark onsets regarding S-W times (a–c) and epoch numbers (d–f) in LD1:1 (n = 5) condition. S-W data were averaged in consecutive 6-min long bins using the data of all light switches during the 24-h recordings. Two bins were analyzed in the pre-switch period (dark hour 48–54 min and 54–60 min) and two bins for the post-switch period (bright hour 0–6 min and 6–12 min). a and d: wakefulness; panel b and e: slow wave sleep; c and f: REM sleep. Significance was tested with two-way ANOVA with time and treatment as factors, followed by Sidak's multiple comparisons test. Differences were checked between the different 6-min time bins before and after the light switch during baseline and after SD and between the baseline vs. after SD values regarding the same time period. Significance levels: *—p < 0.05; **—p < 0.01; ***—p < 0.001. Data are expressed as mean ± S.E.M

Fig. 6

Number of slow waves during baseline and after SD during LD12:12 (n = 6) and LD1:1 (n = 5) conditions. Top row: baseline data, bottom row: sleep deprivation data. a and e: hourly number of SWs during baseline and after SD in LD1:1 and LD12:12 conditions. b and f: slow wave density (number of SWs/SWS time/min) during baseline and after SD in LD1:1 and LD12:12 conditions. Panel c: averaged SW numbers/hour in odd and even hours (LD12:12) and in light and dark hours (LD1:1). d: SW numbers/12 h in the CT1-CT12 and CT13-CT24 periods, in LD1:1 and LD12:12 conditions. g: SW numbers in the first third of the recovery period (R1–R6 hours) SW numbers during CT9, CT11, CT13 baseline hours vs recovery period hours R1, R3, R5 (bright hours in LD1:1). SW numbers during CT8, CT10, CT12 baseline hours vs recovery period hours R2, R4, R6 (dark hours in LD1:1). h: SW numbers summarized in different time periods after SD compared to the corresponding baseline periods without SD. Black and white bars at the x axis show the illumination pattern in the different hours. Significance was tested with two-way ANOVA with time and treatment as factors, followed by Sidak's multiple comparisons test. Significance levels: *, #—p < 0.05; **, ##- p < 0.01; ***, ###—p < 0.001. Data are expressed as mean ± S.E.M

Fig. 7

a: Overview of the lighting conditions and the nomenclature used for the different hours and longer time periods used by this study. Baseline hours named as CT1–CT24 for both lighting conditions (LD12:12 and LD1:1). 6-h total sleep deprivation hours are referenced as SD1–SD6 while the subsequent recovery period is consisted of hours R1–R18. b 10-s representative LFP sample data (low-pass filtered at 128 Hz, top row) from the R1 hour after SD performed in LD12:12 condition. Bottom row shows the corresponding raw multiple unit data filtered between 128 Hz and 4096 Hz. Vertical dotted lines indicate the maxima of the LFP slow waves, the latter represent the DOWN states of the ongoing SCR. On the multiple unit data, the cease of firing is clearly visible simultaneously with the presence of the slow waves. c Parameters of the slow waves analyzed. After the extraction of the slow waves from the raw LFP data, slow waves were averaged for ± 500 ms around the maxima in 1-h long time bins. Then amplitude (A2–A1 in μV), half width (B2–B1 in ms) and slow wave number were summarized in 1-h bins