Shining a light on astrocytes and sleep (Commentary on Pelluru etal.).

Abstract

The idea that glial cells might be important for sleep has a distinguished pedigree. More than 100 years ago, Ramon y Cajal hypothesized that morphological changes in astrocytic coverage of synapses could be the cellular mechanism gating wakefulness and sleep but, after that perhaps prescient insight (Bellesi et al. 2015), the idea languished despite intriguing suggestions and clues over the intervening years. Glial cells, for example, were known to secrete various sleep-promoting substances in vitro which, when introduced in vivo, increased sleep time or indices of sleep intensity. They were also uniquely positioned to sample synaptic and metabolic activity from neurons, and respond in their own ways to create homeostatic control of these processes. Indeed, difficult experiments in anesthetized animals showed that astrocytes buffer extracellular cation concentrations in ways that suggest that these cells influence a classic index of sleep need (EEG slow-wave activity; SWA) (reviewed by Frank, 2013). Direct evidence that glial cells really influence sleep in vivo had to wait for a study by Halassa et al. (2009). This study showed that preventing a form of glial chemical exocytosis (‘gliotransmission’) attenuated signs of sleep drive in vivo. Since that study, there has been a resurgence of interest in the role of glia and sleep. A pubmed search with ‘glial’ and ‘sleep’ in the title or abstract finds only a single study in the decade before 2009 and 59 after 2009. This raises two important questions. What does the future hold, and what have we learned? The future will include the creation of new tools to probe astrocytic function in vivo. Optogenetics is widely used to probe and manipulate neuronal circuits. It involves the expression of light-sensitive ion channels that, when stimulated by specific wavelengths of light, open the channel pore. This has proven to be a highly precise and physiologically relevant tool in neurons because neurons are excitable cells that normally operate via changes in membrane voltage. Several years ago, Deisseroth and colleagues, using an astrocyte-specific viral vector, showed that channelrhodopsin (which passes cations across the membrane) could be expressed in astrocytes in vivo. Upon light stimulation, the pore opened and gliotransmission resulted (Gradinaru et al., 2009). Thus, optogenetics might provide a novel way to interrogate glial function in sleep. This is precisely what Pelluru et al. (2015), decided to do in their study. They expressed channelrhodopsin in astrocytes in a regionally specific manner. The area they selected was the posterior hypothalamus, in astrocytes that envelop histaminergic neurons. These neurons are known to promote wakefulness; thus the idea being tested was that activation of astrocytes would dampen the activity of these neurons and increase sleep or sleep drive, and that is what they found. After verifying that viral expression of channelrhodopsin was limited to astrocytes, light stimulation increased sleep time and also non-rapid-eye movement SWA. These effects were transient, as might be expected if they were physiological as opposed to pathological. This then raises the second question: what have we learned? What we have learned is that astrocytes are capable of influencing sleep under these conditions. What we don't know is whether they do so normally. The channelrhodopsin pore is large and does not normally exist in astrocytes. Opening such pores is akin to suddenly and massively perforating the astrocytic plasma membrane, allowing a rush of cations that may never occur normally (given the normal membrane potential of glial cells). This certainly can result in gliotransmission, but possibly a number of other changes that, perhaps not surprisingly, alter surrounding neuronal activity. The use of optogenetics in astrocytes, therefore, highlights a challenge to the field: creating tools designed for glia based on glial, not neuronal, signaling and biology. Nevertheless, science progresses by such simple elemental steps. The study by Pelluru et al. is important because it pushes the envelope in what we can do to understand the role of glia and sleep. Their results clearly show what might be true. Now, collectively, we must find out whether it is.