The suprachiasmatic nucleus entrains, but does not sustain, circadian rhythmicity in the olfactory bulb

Abstract

The suprachiasmatic nucleus (SCN) of the hypothalamus has been termed the master circadian pacemaker of mammals. Recent discoveries of damped circadian oscillators in other tissues have led to the hypothesis that the SCN synchronizes and sustains daily rhythms in these tissues. We studied the effects of constant lighting (LL) and of SCN lesions on behavioral rhythmicity and Period 1 (Per1) gene activity in the SCN and olfactory bulb (OB). We found that LL had similar effects on cyclic locomotor and feeding behaviors and Per1 expression in the SCN but had no effect on rhythmic Period 1 expression in the OB. LL lengthened the period of locomotor and SCN rhythms by approximately 1.6 hr. After 2 weeks in LL, nearly 35% of rats lost behavioral rhythmicity. Of these, 90% showed no rhythm in Per1-driven expression in their SCN. Returning the animals to constant darkness rapidly restored their daily cycles of running wheel activity and gene expression in the SCN. In contrast, the OB remained rhythmic with no significant change in period, even when cultured from animals that had been behaviorally arrhythmic for 1 month. Similarly, we found that lesions of the SCN abolished circadian rhythms in behavior but not in the OB. Together, these results suggest that LL causes the SCN to lose circadian rhythmicity and its ability to coordinate daily locomotor and feeding rhythms. The SCN, however, is not required to sustain all rhythms because the OB continues to oscillate in vivo when the SCN is arrhythmic or ablated.

Figure 1.

Constant light lengthened and then abolished circadian rhythms in behavior and Per1-luc expression in the SCN but had no effect on rhythms in the OB. Ablation of the SCN also abolished circadian behavior but had no effect on the OB. Representative recordings show wheel running (left), SCN (middle), and OB Per1-luc (right) activity from rats maintained in LD (a-c), LL (d-o), or DD (p, q). After 2 weeks in LL, some rats lost locomotor and SCN rhythmicity, but their OB continued to express rhythmicity (g-l), even after >30 d of behavioral arrhythmicity (m-o). Similarly, the OB remained rhythmic when removed from SCN-lesioned rats, even after 21 d of behavioral arrhythmicity (p, q). The times of light (open bars) and dark (filled bars) during behavioral recordings are shown above the plots in the left column. Animals were killed (^*), and tissues were explanted for bioluminescence recording either 1 hr before the onset of darkness (open and filled bars under b and c show previous light and dark cycle) or ∼1 hr before the predicted onset of daily locomotor activity.

Figure 2.

The period of the SCN, but not the OB, depended on lighting conditions and change with the behavioral period. LL lengthened the period of locomotor activity (25.6 ± 0.1 hr; mean ± SEM) and of SCN bioluminescence rhythms (26.2 ± 0.3 hr) by ∼1.6 hr from LD entrained conditions. When animals were switched from LL to DD, locomotor activity (24.6 ± 0.06 hr) and Per1-luc (24.4 ± 0.17 hr) periods significantly shortened. In contrast, the period of the cultured OB did not significantly change after any of the illumination conditions (22.4 ± 0.7 hr for LD; 22.8 ± 1.6 hr for LL; 21.6 ± 0.5 hr for DD after LL; comparison of LD vs LL vs DD locomotor activity, F_(2,32) = 132, p < 0.001; SCN Per1-luc activity, F_(2,24) = 7.07, p < 0.005; OB Per1-luc activity, F_(2,12) = 0.15, p = 0.86; one-way ANOVA and Tukey's test).

Figure 3.

The time of peak Per1-luc expression in SCN and OB cultures was tightly regulated in LD but disrupted by LL. The time of the first peak of each explant in bioluminescence provided a phase marker for in vivo rhythmicity relative to the time of light onset in the animal colony (ZT 0), activity onset (CT 12), or projected activity onset had they remained behaviorally rhythmic (CT 12). a, b, SCN cultures from animals kept in LD cycles peaked ∼9.5 hr after the time of projected dawn (ZT 9.5 hr; p = 0.001; r = 0.98; Rayleigh test), whereas OB cultures peaked ∼2.5 hr later (ZT 12.0 hr; p = 0.001; r = 0.98). c, d, SCN cultures from rhythmic animals kept in LL did not show significant phase clustering (p = 0.1; r = 0.57). OB cultures from the same animals, however, peaked at CT 10.7 hr (p = 0.006; r = 0.8). e, f, Of the subset of animals that became behaviorally arrhythmic in LL, only one had a rhythmic SCN in vitro, although all had rhythmic OBs that peaked at variable times (p = 0.7; r = 0.3). g, Similarly, the OB of animals rendered behaviorally arrhythmic by SCN ablation also showed a broadened distribution at peak times (p = 0.47; r = 0.4). The times of the previous light (open bars) and dark (filled bars) schedule are shown above each plot.

Figure 4.

The normal phase relationship between the SCN and OB is disrupted by constant light. The phase relationship between SCN and OB explants was calculated by subtracting the time of the first peak in Per1-luc expression of the SCN from that of the OB. Positive values thus indicate when the rhythm in the SCN preceded that in the OB from the same animal. Constant light abolished the reliable lag from SCN to OB seen in LD. Lag was statistically grouped at 2.5 hr (p = 0.01; r = 0.97) for LD, whereas the LL group was not grouped (p = 0.06; r = 0.6).