Division of Sleep Medicine, Brigham & Women's Hospital, Harvard Medical School, Boston, MA 02115, USA.
J Biol Rhythms. 2011 Oct;26(5):441-53. doi: 10.1177/0748730411414163.
Early attempts to characterize free-running human circadian rhythms generated three notable results: 1) observed circadian periods of 25 hours (considerably longer than the now established 24.1- to 24.2-hour average intrinsic circadian period) with sleep delayed to later circadian phases than during entrainment; 2) spontaneous internal desynchrony of circadian rhythms and sleep/wake cycles--the former with an approximately 24.9-hour period, and the latter with a longer (28-68 hour) or shorter (12-20 hour) period; and 3) bicircadian (48-50 hour) sleep/wake cycles. All three results are reproduced by Kronauer et al.'s (1982) coupled oscillator model, but the physiological basis for that phenomenological model is unclear. We use a physiologically based model of hypothalamic and brain stem nuclei to investigate alternative physiological mechanisms that could underlie internal desynchrony. We demonstrate that experimental observations can be reproduced by changes in two pathways: promotion of orexinergic (Orx) wake signals, and attenuation of the circadian signal reaching hypothalamic nuclei. We reason that delayed sleep is indicative of an additional wake-promoting drive, which may be of behavioral origin, associated with removal of daily schedules and instructions given to participants. We model this by increasing Orx tone during wake, which reproduces the observed period lengthening and delayed sleep. Weakening circadian input to the ventrolateral preoptic nucleus (possibly mediated by the dorsomedial hypothalamus) causes desynchrony, with observed sleep/wake cycle period determined by degree of Orx up-regulation. During desynchrony, sleep/wake cycles are driven by sleep homeostasis, yet sleep bout length maintains circadian phase dependence. The model predicts sleep episodes are shortest when started near the temperature minimum, consistent with experimental findings. The model also correctly predicts that it is possible to transition to bicircadian rhythms from either a synchronized or desynchronized state. Our findings suggest that feedback from behavioral choices to physiology could play an important role in spontaneous internal desynchrony.
1）观察到的昼夜周期为 25 小时（明显长于现在确立的 24.1 到 24.2 小时的内在昼夜周期平均时间），睡眠会延迟到比适应期更晚的昼夜阶段；2）昼夜节律和睡眠/觉醒周期的自发内部失同步——前者的周期约为 24.9 小时，后者的周期较长（28-68 小时）或较短（12-20 小时）；3）双昼夜（48-50 小时）睡眠/觉醒周期。所有这三个结果都被 Kronauer 等人（1982 年）的耦合振荡器模型所复制，但该现象模型的生理基础尚不清楚。我们使用下丘脑和脑干核的基于生理的模型来研究可能构成内部失同步基础的替代生理机制。我们证明，通过改变两种途径可以再现实验观察结果：促进食欲素（Orx）觉醒信号和减弱到达下丘脑核的昼夜信号。我们推断，睡眠延迟表明存在额外的促觉醒驱动力，这可能是行为起源的，与日常时间表的去除和给予参与者的指令有关。我们通过在觉醒期间增加 Orx 音来模拟这种情况，这复制了观察到的周期延长和睡眠延迟。减弱对腹外侧视前核的昼夜输入（可能由下丘脑背内侧核介导）会导致失同步，观察到的睡眠/觉醒周期的周期由 Orx 上调的程度决定。在失同步期间，睡眠/觉醒周期由睡眠稳态驱动，但睡眠爆发长度保持昼夜相位依赖性。该模型预测，当在体温最低附近开始时，睡眠发作最短，这与实验结果一致。该模型还正确预测，从同步或失同步状态过渡到双昼夜节律是可能的。我们的研究结果表明，来自行为选择的反馈对生理学可能在自发内部失同步中发挥重要作用。
