Phase shifts in circadian peripheral clocks caused by exercise are dependent on the feeding schedule in PER2::LUC mice

Circadian rhythms are regulated by the suprachiasmatic nucleus (SCN) clock, which is the main oscillator and peripheral clock. SCN clock can be entrained by both photic and non-photic stimuli, and an interaction exists between photic and non-photic entrainment. Moreover, peripheral circadian clocks can be entrained not only by scheduled restricted feeding, but also by scheduled exercise. Thus, the entrainment of peripheral circadian clocks may be the result of an interaction between the entrainment caused by feeding and exercise. In this study, we examined the effect of wheel-running exercise on the phase of the peripheral clocks (kidney, liver and submandibular gland) in PER2::LUC mice under various feeding schedules. Phase and waveforms of the peripheral clocks were not affected by voluntary wheel-running exercise. Exercise for a period of 4 h during the early dark period (morning) delayed the peripheral clocks, while exercise for the same duration during the late dark period (evening) advanced the peripheral clocks. The feeding phase was advanced and delayed by evening and morning exercise, respectively, suggesting that the feeding pattern elicited by the scheduled exercise may entrain the peripheral clocks. Exercise did not affect the phase of the peripheral clock under the 1 meal per day schedule. When the phase of the peripheral clocks was advanced by the feeding schedule of 2 or 4 meals per day during light and/or dark periods, wheel-running exercise during the morning period significantly and equally shifted the phase of all organs back to the original positions observed in mice maintained under free-feeding conditions and with no exercise. When the schedule of 2 meals per day during the dark period failed to affect the phase of peripheral clock, morning exercise did not affect the phase. Wheel-running exercise increased the levels of serum corticosterone, and the injection of dexamethasone/corticosterone instead of exercise shifted a phase that had advanced under the feeding schedule of 2 meals per day, back to the normal position. The liver and submandibular glands exhibit higher sensitivity to dexamethasone than the kidneys. In adrenalectomized mice, treadmill-induced normalization of the advanced phase under a feeding schedule of 2 meals per day was not observed. In summary, scheduled exercise-induced phase shifts were weaker compared to scheduled feeding-induced phase shifts. The phase advance caused by the feeding schedule of 2 or 4 meals per day was suppressed by wheel-running, treadmill exercise or dexamethasone/corticosterone injection in the early dark period (morning). Corticosterone release may be involved in exercise-induced phase shift of peripheral clocks. These results suggest that there is an interaction between the phase shifts caused by feeding and exercise schedules in peripheral clocks.