Differentiating external zeitgeber impact on peripheral circadian clock resetting

Abstract

Circadian clocks regulate physiological functions, including energy metabolism, along the 24-hour day cycle. The mammalian clock system is organized in a hierarchical manner with a coordinating pacemaker residing in the hypothalamic suprachiasmatic nucleus (SCN). The SCN clock is reset primarily by the external light-dark cycle while other zeitgebers such as the timing of food intake are potent synchronizers of many peripheral tissue clocks. Under conflicting zeitgeber conditions, e.g. during shift work, phase synchrony across the clock network is disrupted promoting the development of metabolic disorders. We established a zeitgeber desynchrony (ZD) paradigm to quantify the differential contributions of the two main zeitgebers, light and food, to the resetting of specific tissue clocks and the effect on metabolic homeostasis in mice. Under 28-hour light-dark and 24-hour feeding-fasting conditions SCN and peripheral clock, as well as activity and hormonal rhythms showed specific periodicities aligning in-between those of the two zeitgebers. During ZD, metabolic homeostasis was cyclic with mice gaining weight under synchronous and losing weight under conflicting zeitgeber conditions. In summary, our study establishes an experimental paradigm to compare zeitgeber input in vivo and study the physiological consequences of chronodisruption.

Figure 1

Locomotor activity under zeitgeber desynchrony (ZD) conditions. (a) Schematic representation of the ZD paradigm. Rectangles indicate in-phase (food access during the dark phase; solid frame) and anti-phase days (food access during the light phase; dashed frame). (b) Fractions of animals that completely adapted to the 28-hour LD cycle (period = 28 h; n = 8) and of free-running animals (period < 28 h; n = 53) under ZD conditions. (c) Distribution of dominant free-running periods (main periodogram peaks) under ZD conditions. The dotted curve shows a Lorentzian distribution fit (peak at 25.6 h). (d) Representative activity recording of one mouse over the course of 5 ZD cycles. Green lines depict different free-running rhythm components as indicated in (e). (e) χ² periodogram analysis of running-wheel activity of the mouse shown in (d). Green numbers indicate free-run period components as depicted in (d). (f,g) Daily activity onset relative to feeding intervals throughout the experiment (f) and averaged activity onsets over one FD-cycle relative to the start of food access (g) during ZD. Three mice were excluded due to highly fragmented activity patterns. Data are shown as means ± SEM, n = 58. Dark phases are depicted by dark grey, food access by light grey shading.

Figure 2

Activity profiles on ZD in-phase and anti-phase days. (a) Mean activity profiles at in-phase (solid black line) and anti-phase (dashed grey line) days relative to food access (180° = “food in”). (b) Feeding/fasting phase activity distribution at in-phase (black columns) and anti-phase days (grey columns; ****p < 0.0001 between days, ^####p < 0.0001 between feeding and fasting phase on the same day, two-way ANOVA). (c) Total running-wheel activity at in-phase (black) and anti-phase days (grey; ****p < 0.0001, two-tailed, paired t-test). Values are means (±SEM in b,c), n = 60. One animal was excluded due to missing values in the activity recording of respective days.

Figure 3

Regulation of SCN clock gene expression under ZD conditions. (a) Diurnal mRNA expression profiles for Bmal1 (upper panel) and Per2 (lower panel) on in-phase (solid black line) and anti-phase (dashed grey line) days determined by ³⁵S-UTP in situ hybridisation. Data represent 3 independent experiments. Grey shading depicts time of food access (180° = “food in”). (b) Phase shifts of SCN clock gene expression rhythms between in-phase and anti-phase days. Data are shown as means ± SEM, n = 2–3 animals per time point.

Figure 4

Regulation of clock gene expression in peripheral tissues under ZD conditions. (a) Diurnal mRNA expression profiles for Bmal1 (top), Per2 (middle), and Dbp (bottom) in three peripheral tissues on in-phase (solid black line) and anti-phase (dashed grey line) days determined by qPCR. Data represent 3 independent experiments. Grey shading depicts time of food access (180° = “food in”). (b) Phase shifts of peripheral tissue clock gene expression rhythms between in-phase and anti-phase days. Data are shown as means ± SEM, n = 2–4 animals per time point.

Figure 5

Corticosterone and leptin regulation under ZD conditions. (a) Diurnal serum corticosterone (a) and leptin (b) profiles on in-phase (black solid lines) and anti-phase days (grey dashed lines). Data represent 3 independent experiments. Grey shading depicts time of food access (180° = “food in”). (c) Phase shifts of corticosterone and leptin serum rhythms between in-phase and anti-phase conditions. Data are shown as means ± SEM, n = 2–4 animals per time point.

Figure 6

Changes in food intake and body weight under ZD conditions. (a) Food intake (closed squares) and body weight gain (open squares) throughout 5 weeks of the ZD paradigm. In-phase (solid black lines) and anti-phase days (dashed grey lines) are indicated. (b) Normalized food intake (closed squares) and weight change (open squares) profiles across the ZD cycle. (c) Normalized food consumption, (d) weight gain, and (e) energy conversion on in-phase and anti-phase days. All data are shown as means ± SEM (n = 45, ****p < 0.0001, two-tailed, paired t-tests). (f) Glucose tolerance test over time (left) and calculated area under the curve (right) on in-phase (n = 4) and anti-phase (n = 6) days. Due to technical problems two animals in the in-phase condition had to be excluded. Data are shown as means ± SEM, ***p < 0.001, condition effect ^§§p < 0.005, time effect ^$$$$ p < 0.0001, two-way ANOVA.

Figure 7

Determination of zeitgeber impact on different circadian rhythms under ZD conditions. (a) Phase shifts between ZD in-phase and anti-phase days for different endocrine, tissue clock, and behavioural rhythms conditions (means ± SEM, n = 3 for tissues/hormones, n = 58 for activity, one-sample t tests were used to determine statistical differences to the two zeitgeber rhythms (0 = feeding; ± 12 = light; ^*p < 0.05, ^**p < 0.005, ^****p < 0.0001 vs. light, ^#p < 0.05, ^####p < 0.005 vs. food). (b) Model of circadian rhythm adaptation under conflicting zeitgeber (ZD) conditions with phase coherence on in-phase and maximal misalignment on anti-phase days.