Friedrich K. Stephan

The "other" circadian system: food as a Zeitgeber.

It is not surprising that limiting food access to a particular time of day has profound effects on the behavior and physiology of animals. It has been clear for some time that pre-meal behavioral activation, a rise in core temperature, elevated serum corticosterone, and an increase in duodenal disaccharidases are under circadian control and that the observed circadian properties are not abolished by lesions of the suprachiasmatic nucleus (SCN), but the search for the locus of a separate food-entrainable oscillator (FEO) has not been successful. The cloning of circadian clock genes and the discovery that these genes are expressed in many central nervous system structures outside the SCN and in peripheral tissues have led to new strategies for investigating potential loci of an FEO. Recent findings concerning the entrainment of clock gene expression in the central nervous system and in peripheral tissues by periodic food access are presented, and the implications of these findings for a better understanding of a circadian system that entrains to meals, rather than to light, are discussed.

Anticipation of 24-hr feeding schedules in rats with lesions of the suprachiasmatic nucleus.

Anticipatory wheel running in response to resticted feeding schedules was studied in rats with lesions of the suprachiasmatic nucleus (SCN) and in sham-operated controls. Despite the absence of circadian periodicity in free-feeding conditions, rats with SCN lesions anticipated restricted access to food at 24-hr intervals in the presence of a light-dark cycle and in constant light. Neither rats with SCN lesions nor controls were able to anticipate feedings at 18-hr intervals. Adrenalectomy did not prevent anticipatory activity to a 24-hr feeding schedule in either group. These results suggest that circadian oscillators outside SCN can be entrained by restricted feeding schedules or, alternatively, that anticipatory activity is based on the hourglass principle, i.e., a clock which requires daily resetting.

Entrainment of circadian rhythms by feeding schedules in rats with suprachiasmatic lesions.

Rats with lesions of the suprachiasmatic nucleus (SCN) and controls were maintained in constant light and exposed to a restricted feeding schedule at 23- and 24-hr intervals, as well as to a 12-hr phase shift in the feeding schedule. Despite the absence of circadian periodicity in activity or drinking in ad lib. conditions, rats with SCN lesions showed anticipatory wheel running to both feeding schedules, comparable to sham-operated rats. Following the 12-hr phase shift, transients qualitatively similar to those seen following phase shifts in the light—dark cycle were observed. During a 3-day period of total food deprivation following prolonged entrainment to a 24-hr feeding schedule, wheel running persisted with a near 24-hr periodicity while return to ad lib. conditions resulted in a rapid desynchronization of activity. These results indicate that anticipatory wheel running in rats with SCN lesions is based on endogenous circadian oscillators which are entrainable by feeding schedules in the circadian range. Apparently such oscillators free run under certain conditions (food deprivation) but become rapidly desynchronized in others (ad lib. feeding). The evidence strongly supports a multioscillator model of the circadian system in mammals.

Limits of Entrainment to Periodic Feeding in Rats with Suprachiasmatic Lesions

Summary1.Rats with lesions of the suprachiasmatic nucleus were maintained in constant darkness and exposed to restricted feeding at periods of 20, 21, 22, 23, 24, 27, 29, 31, and 33 h. The limits of entrainment of activity in anticipation of food availability were studied.2.Twelve out of 16 rats failed to show a circadian rhythm in activity in ad lib. feeding conditions.3.Almost all rats entrained to the 24 h feeding schedule. A change from 24 to 22 or to 31 h did not consistently produce entrainment. However, a gradual increase in the period of the schedule resulted in some entrainment to the 31 and even 33 h schedule. All rats except one entrained to schedules between 23 and 29 h.4.Near the limits of entrainment, activity rhythms often free ran at periods deviating considerably from the period of the feeding schedule.5.In ad lib. conditions, after prolonged exposure to feeding schedules, activity rhythms persisted only in a few cases. However, during total food deprivation activity rhythms continued for at least 2 cycles.6.These results indicate that circadian pacemakers outside the suprachiasmatic nucleus are entrainable by feeding, and that there are some similarities between the entrainment of circadian rhythms by light-dark cycles and by food.

Elimination of circadian rhythms in drinking, activity, sleep, and temperature by isolation of the suprachiasmatic nuclei.

Isolation or destruction of the suprachiasmatic nuclei of rats resulted in an immediate and apparently permanent elimination of circadian rhythms in drinking, wheel running, slow wave sleep, cortical arousal, and brain temperature. While activity levels were reduced, total daily water intakes and amounts of sleep and arousal were not substantially affected. Paradoxical sleep was arhythmic in some animals, whereas others showed a predominantly nocturnal pattern following the lesions. The results support the hypothesis that the suprachiasmatic nuclei may serve as a central neural pacemaker responsible for the generation and entrainment of a number of circadian rhythms in physiology and behavior, and that the regulation of rhythmic functions is mediated by neural efferents from the suprachiasmatic nuclei.

Phase shifts of circadian rhythms in activity entrained to food access

Rats anticipate daily 2 hr meals with a sharp increase in activity several hours prior to food availability. The present experiment examined the response to phase shifts of food access in rats with lesions of the suprachiasmatic nuclei (SCN). Following entrainment of activity to 2 hr of food per day, food access was phase delayed or phase advanced by 4, 6, or 8 hr. All rats responded to phase delays of 4 or 6 hr with an increase in the duration of anticipatory activity so that transients appeared mostly in activity onset. Following 8 hr phase delays, clear delaying transients in both activity onset and end were observed. Only a few rats showed advancing transients in activity after phase advances of food access. In response to 6 hr and 8 hr phase advances, 3 different responses occurred: (a) activity re-entrained to food access by the 2nd or 3rd day without clear intervening transients, (b) activity phase shifted by means of distinct delaying transients and (c) delaying transients occurred in one component of activity while a second component of activity appeared at the new phase position by the second or third day. These results provide further evidence that anticipation of food access is mediated by a circadian mechanism which is functionally independent of the SCN and illustrate some similarities as well as considerable differences between circadian rhythms entrained by feeding and those entrained by light-dark cycles.

Glucose, but not fat, phase shifts the feeding-entrained circadian clock

To study the ability of single macronutrients to entrain or phase shift the feeding entrainable circadian oscillator, rats with lesions of the suprachiasmatic nucleus were first maintained on a single daily meal of lab chow until robust anticipatory approaches to the feeder or anticipatory wheel running was established. The meal time was then delayed by 8 h and chow was replaced with a 25-mL solution of 0.2% saccharin or 25 mL of saccharin plus 15 g of glucose. For other phase shifts, rats received either 6 mL of vegetable oil or mineral oil for 2 consecutive days. Consumption of about 6 g (24 kcal) or more of glucose resulted in robust delaying transients on the days after ingestion, whereas saccharin induced only small delays consistent with the initiation of a free-running rhythm with a period greater than 24 h. Surprisingly, consumption of 5.5 g of vegetable oil (47 kcal) did not result in delays greater than those in rats receiving mineral oil. The introduction of oil also produced a severe reduction in approaches to the feeder which could be alleviated by placing inaccessible chow in the feeders between oil meals. Phase shifts with oil were repeated with rats housed in wheels using anticipatory wheel running as a phase marker to assess whether the lack of phase shifts with fat was apparatus dependent. As was the case with approach behavior, anticipatory wheel running was not significantly delayed by vegetable oil consumption. These results indicate that a simple monosaccharide, glucose, has zeitgeber properties for the feeding entrainable oscillator. Vegetable oil, despite a higher caloric content, may be ineffective because of slower gastric emptying and nutrient absorption or because fat is not a good zeitgeber for the feeding entrained circadian oscillator.

The role of period and phase in interactions between feeding- and light-entrainable circadian rhythms

Sixteen blind male rats were maintained on ad lib food and water for 115 days to obtain stable free-running rhythms. Fifteen rats were then exposed to restricted feeding (RF, 4 hr/cycle) for 52 days and the period of food access (T) differed from the period of the free-running rhythm (tau) by 0.05 to 0.30 hr. Among 15 animals exposed to RF, deceleration of the free-running rhythm was observed in 11 rats, small accelerations occurred in 2 rats, and tau remained unchanged in 2 rats exposed to RF and one ad lib rat. The free-running rhythm assumed the period of RF in 3 rats where T-tau was less than 0.1 hr and the change in tau persisted for 42 days after RF. However, other rats failed to synchronize their free-running rhythm despite small period differences. Although these results show that the two underlying circadian pacemaking systems are not functionally independent, they appear to be only weakly coupled.

Circadian rhythm dissociation induced by periodic feeding in rats with suprachiasmatic lesions

Rats with lesions of the suprachiasmatic nucleus (SCN) were maintained in constant darkness and placed on restricted feeding schedules with 1 h access to food twice per day. When both schedules had a 24 h period and food access was spaced 12 h apart, all rats with SCN lesions displayed anticipatory activity to both feedings. In the first experiment, the period of one schedule was then changed to 25 h while the other continued with a period of 24 h. None of the rats with SCN lesions were able to anticipate both schedules simultaneously. In a second experiment, 24 h and 24.5 h schedules were used. Five of 6 rats with SCN lesions anticipated both schedules for at least 9 days and 2 of these anticipated both schedules for 48 days. This forced dissociation of activity into two components with different periods is consistent with the hypothesis that entrainment of activity by restricted food access is mediated by more than one circadian pacemaker.

The effects of hypothalamic knife cuts on drinking rhythms and the estrus cycle of the rat

Knife cuts placed posterior, anterior, or lateral to the suprachiasmatic nuclei failed to eliminate the drinking and eating rhythms of rats. Partial isolation of the suprachiasmatic nuclei, interrupting simultaneously lateral, caudal, and dorsal efferents, permanently abolished drinking rhythms in the presence of a light—dark cycle and in constant dim light. This manipulation also eliminated the estrus cycle of rats. The results suggest that drinking rhythms are controlled by neural efferents from the suprachiasmatic nuclei. While the exact course of these efferents could not be determined, they appear to be anatomically different from those regulating hormonal rhythms of the hypothalamo-hypophyseal system.