Supplemental Digital Content is available in the text.

The basal forebrain (BF) plays a crucial role in cortical activation. Our previous study showed that activation of cholinergic BF neurons alone is sufficient to suppress slow-wave sleep (SWS) and promote wakefulness and rapid-eye-movement (REM) sleep. However, the exact role of silencing cholinergic BF neurons in the sleep-wake cycle remains unclear. We inhibitied the cholinergic BF neurons genetically targeted with archaerhodopsin (Arch) with yellow light to clarify the role of cholinergic BF neurons in the sleep-wake cycle. Bilateral inactivation of cholinergic BF neurons genetically targeted with archaerhodopsin prolonged SWS and decreased the probability of awakening from SWS in mice. However, silencing these neurons changed neither the duration of wakefulness or REM sleep, nor the probability of transitions to other sleep-wake episodes from wakefulness or REM sleep. Furthermore, silencing these neurons for 6 h within the inactive or active period increased the duration of SWS at the expense of the duration of wakefulness, as well as increasing the number of prolonged SWS episodes (120-240 s). The lost wakefulness was compensated by a delayed increase of wakefulness, so the total duration of SWS and wakefulness during 24 h was kept stable. Our results indicate that the main effect of these neurons is to terminate SWS, whereas wakefulness or REM sleep may be determined by co-operation of the cholinergic BF neurons with other arousal-sleep control systems.

Although extensively studied, the exact role of sleep in learning and memory is still not very clear. Sleep deprivation has been most frequently used to explore the effects of sleep on learning and memory, but the results from such studies are inevitably complicated by concurrent stress and distress. Furthermore, it is not clear whether there is a strict time-window between sleep and memory consolidation. In the present study we were able to induce time-locked slow-wave sleep (SWS) in mice by optogenetically stimulating GABAergic neurons in the parafacial zone (PZ), providing a direct approach to analyze the influences of SWS on learning and memory with precise time-windows. We found that SWS induced by light for 30 min immediately or 15 min after the training phase of the object-in-place task significantly prolonged the memory from 30 min to 6 h. However, induction of SWS 30 min after the training phase did not improve memory, suggesting a critical time-window between the induction of a brief episode of SWS and learning for memory consolidation. Application of a gentle touch to the mice during light stimulation to prevent SWS induction also failed to improve memory, indicating the specific role of SWS, but not the activation of PZ GABAergic neurons itself, in memory consolidation. Similar influences of light-induced SWS on memory consolidation also occurred for Y-maze spatial memory and contextual fear memory, but not for cued fear memory. SWS induction immediately before the test phase had no effect on memory performance, indicating that SWS does not affect memory retrieval. Thus, by induction of a brief-episode SWS we have revealed a critical time window for the consolidation of hippocampus-dependent memory.

The GABAergic neurons in the parafacial zone (PZ) play an important role in sleep-wake regulation and have been identified as part of a sleep-promoting center in the brainstem, but the long-range connections mediating this function remain poorly characterized. Here, we performed whole-brain mapping of both the inputs and outputs of the GABAergic neurons in the PZ of the mouse brain. We used the modified rabies virus EnvA-ΔG-DsRed combined with a Cre/loxP gene-expression strategy to map the direct monosynaptic inputs to the GABAergic neurons in the PZ, and found that they receive inputs mainly from the hypothalamic area, zona incerta, and parasubthalamic nucleus in the hypothalamus; the substantia nigra, pars reticulata and deep mesencephalic nucleus in the midbrain; and the intermediate reticular nucleus and medial vestibular nucleus (parvocellular part) in the pons and medulla. We also mapped the axonal projections of the PZ GABAergic neurons with adeno-associated virus, and defined the reciprocal connections of the PZ GABAergic neurons with their input and output nuclei. The newly-found inputs and outputs of the PZ were also listed compared with the literature. This cell-type-specific neuronal whole-brain mapping of the PZ GABAergic neurons may reveal the circuits underlying various functions such as sleep-wake regulation.