Raw data underlying all Figs.

Sleep is a behavior that is conserved throughout the animal kingdom. Yet, despite extensive studies in humans and animal models, the exact function or functions of sleep remain(s) unknown. A complicating factor in trying to elucidate the function of sleep is the complexity and multiplicity of neuronal circuits that are involved in sleep regulation. It is conceivable that distinct sleep-regulating circuits are only involved in specific aspects of sleep and may underlie different sleep functions. Thus, it would be beneficial to assess the contribution of individual circuits in sleep’s putative functions. The intricacy of the mammalian brain makes this task extremely difficult. However, the fruit fly Drosophila melanogaster, with its simpler brain organization, available connectomics, and unparalleled genetics, offers the opportunity to interrogate individual sleep-regulating centers. In Drosophila, neurons projecting to the dorsal fan-shaped body (dFB) have been proposed to be key regulators of sleep, particularly sleep homeostasis. We recently demonstrated that the most widely used genetic tool to manipulate dFB neurons, the 23E10-GAL4 driver, expresses in 2 sleep-regulating neurons (VNC-SP neurons) located in the ventral nerve cord (VNC), the fly analog of the vertebrate spinal cord. Since most data supporting a role for the dFB in sleep regulation have been obtained using 23E10-GAL4, it is unclear whether the sleep phenotypes reported in these studies are caused by dFB neurons or VNC-SP cells. A recent publication replicated our finding that 23E10-GAL4 contains sleep-promoting neurons in the VNC. However, it also proposed that the dFB is not involved in sleep regulation at all, but this suggestion was made using genetic tools that are not dFB-specific and a very mild sleep deprivation protocol. In this study, using a newly created dFB-specific genetic driver line, we demonstrate that optogenetic activation of the majority of 23E10-GAL4 dFB neurons promotes sleep and that these neurons are involved in sleep homeostasis. We also show that dFB neurons require stronger stimulation than VNC-SP cells to promote sleep. In addition, we demonstrate that dFB-induced sleep can consolidate short-term memory (STM) into long-term memory (LTM), suggesting that the benefit of sleep on memory is not circuit-specific. Finally, we show that dFB neurons are neurochemically heterogeneous and can be divided in 3 populations. Most dFB neurons express both glutamate and acetylcholine, while a minority of cells expresses only one of these 2 neurotransmitters. Importantly, dFB neurons do not express GABA, as previously suggested. Using neurotransmitter-specific dFB tools, our data also points at cholinergic dFB neurons as particularly potent at regulating sleep and sleep homeostasis.

睡眠是整个动物界均保守存在的行为。尽管针对人类与多种动物模型已开展了大量研究，但睡眠的确切功能至今仍未明晰。阐明睡眠功能的一大复杂因素，在于睡眠调控所涉及的神经环路兼具复杂性与多样性。可以想见，不同的睡眠调控环路仅参与睡眠的特定环节，并对应不同的睡眠功能。因此，评估单个环路在睡眠推定功能中的贡献将极具研究价值。哺乳动物大脑的复杂性使得这项任务极具挑战性。但果蝇——黑腹果蝇（Drosophila melanogaster）——凭借其更为简单的大脑结构、可获取的连接组数据以及无与伦比的遗传学工具，为研究单个睡眠调控中枢提供了契机。在黑腹果蝇中，投射至背侧扇形体（dorsal fan-shaped body, dFB）的神经元被认为是睡眠的关键调控因子，尤其在睡眠稳态调控中发挥重要作用。我们此前的研究证实，目前最常用于操控dFB神经元的遗传工具——23E10-GAL4驱动系——会在位于腹神经索（ventral nerve cord, VNC，即脊椎动物脊髓的果蝇同源结构）的2个睡眠调控神经元（VNC-SP神经元）中表达。由于绝大多数支持dFB参与睡眠调控的实验数据均通过23E10-GAL4获得，因此目前尚不清楚此类研究中报道的睡眠表型，究竟是由dFB神经元还是VNC-SP细胞所介导。近期一项研究重复了我们关于23E10-GAL4在腹神经索中存在睡眠促进神经元的发现，但该研究同时提出dFB完全不参与睡眠调控——这一结论所使用的遗传工具并非dFB特异性，且采用了极为温和的睡眠剥夺方案。本研究通过全新构建的dFB特异性遗传驱动系，证实了23E10-GAL4标记的大多数dFB神经元经光遗传激活后可促进睡眠，且此类神经元参与睡眠稳态调控。我们还发现，dFB神经元要促进睡眠所需的刺激强度高于VNC-SP细胞。此外，本研究证实，由dFB神经元介导的睡眠可将短时记忆（short-term memory, STM）巩固为长时记忆（long-term memory, LTM），这表明睡眠对记忆的益处并非环路特异性的。最后，我们发现dFB神经元在神经化学层面具有异质性，可分为3个亚群：大多数dFB神经元同时表达谷氨酸与乙酰胆碱，仅少数细胞仅表达这两种神经递质中的一种。值得注意的是，与此前的结论不同，dFB神经元并不表达γ-氨基丁酸（GABA）。借助神经递质特异性的dFB研究工具，我们的数据还显示，胆碱能dFB神经元在调控睡眠及睡眠稳态方面的作用尤为显著。