No abstract

Sleep is regulated by two mechanisms: the homeostatic process and the circadian clock. Light affects sleep and alertness by entraining the circadian clock, and acutely inducing sleep/alertness, in a manner mediated by intrinsically photosensitive retinal ganglion cells. Because intrinsically photosensitive retinal ganglion cells are believed to be minimally sensitive to red light, which is widely used for illumination to reduce the photic disturbance to nocturnal animals during the dark phase. However, the appropriate intensity of the red light is unknown. In the present study, we recorded electroencephalograms and electromyograms of freely moving mice to investigate the effects of red light emitted by light-emitting diodes at different intensities and for different durations on the sleep–wake behavior of mice. White light was used as a control. Unexpectedly, red light exerted potent sleep-inducing effects and changed the sleep architecture in terms of the duration and number of sleep episodes, the stage transition, and the EEG power density when the intensity was >20 lx. Subsequently, we lowered the light intensity and demonstrated that red light at or below 10 lx did not affect sleep–wake behavior. White light markedly induced sleep and disrupted sleep architecture even at an intensity as low as 10 lx. Our findings highlight the importance of limiting the intensity of red light (⩽10 lx) to avoid optical influence in nocturnal behavioral experiments, particularly in the field of sleep and circadian research.

Light regulates daily sleep rhythms by a neural circuit that connects intrinsically photosensitive retinal ganglion cells (ipRGCs) to the circadian pacemaker, the suprachiasmatic nucleus. Light, however, also acutely affects sleep in a circadian-independent manner. The neural circuits involving the acute effect of light on sleep remain unknown. Here we uncovered a neural circuit that drives this acute light response, independent of the suprachiasmatic nucleus, but still through ipRGCs. We show that ipRGCs substantially innervate the preoptic area (POA) to mediate the acute light effect on sleep in mice. Consistently, activation of either the POA projecting ipRGCs or the light-responsive POA neurons increased non-rapid eye movement (NREM) sleep without influencing REM sleep. In addition, inhibition of the light-responsive POA neurons blocked the acute light effects on NREM sleep. The predominant light-responsive POA neurons that receive ipRGC input belong to the corticotropin-releasing hormone subpopulation. Remarkably, the light-responsive POA neurons are inhibitory and project to well-known wakefulness-promoting brain regions, such as the tuberomammillary nucleus and the lateral hypothalamus. Therefore, activation of the ipRGC-POA circuit inhibits arousal brain regions to drive light-induced NREM sleep. Our findings reveal a functional retina-brain circuit that is both necessary and sufficient for the acute effect of light on sleep.

Retinal ganglion cells (RGCs), the output neurons of the retina, allow us to perceive our visual environment. RGCs respond to rod/cone input through the retinal circuitry, however, a small population of RGCs are in addition intrinsically photosensitive (ipRGCs) and project to unique targets in the brain to modulate a broad range of subconscious visual behaviors such as pupil constriction and circadian photoentrainment. Despite the discovery of ipRGCs nearly two decades ago, there is still little information about how or if conventional RGCs (non-ipRGCs) target ipRGC-recipient nuclei to influence subconscious visual behavior. Using a dual recombinase fluorescent reporter strategy, we showed that conventional RGCs innervate many subconscious ipRGC-recipient nuclei, apart from the suprachiasmatic nucleus. We revealed previously unrecognized stratification patterns of retinal innervation from ipRGCs and conventional RGCs in the ventral portion of the lateral geniculate nucleus. Further, we found that the percent innervation of ipRGCs and conventional RGCs across ipsiand contralateral nuclei differ. Our data provide a blueprint to understand how conventional RGCs and ipRGCs innervate different brain regions to influence subconscious visual behaviors.