Astrocytes are star-shaped cells that account for about 40% of cells in the mammalian brain. Initially considered to be the ‘glue’ that sticks neurons together, astrocytes actually have crucial roles in brain homeostasis and in regulating the formation, maturation, function and elimination of synapses, the connections through which neurons communicate with each other1–4. Although much progress has been made in elucidating the roles of astrocytes1–4, our understanding of how they regulate neural circuits and affect behaviours that are associated with neurological and psychiatric disorders is just emerging5–9. Writing in Cell, Nagai et al.10 present evidence in mice that selective activation of astrocytes in the striatum, a brain region that integrates signals from many parts of the brain to coordinate voluntary movement11, drives behavioural changes that resemble the symptoms of attention-deficit hyperactivity disorder (ADHD) in humans through a dialogue with striatal neurons.
ADHD is a prevalent psychiatric and neurodevelopmental disorder that affects approximately 5% of children worldwide, and its major symptoms include excessive activity (or restlessness) and difficulty in sustaining attention12. Although dysfunction in the striatum has been implicated in ADHD13, the underlying mechanisms of how the striatum — and, in particular, striatal astrocytes — might contribute to the disorder, remain elusive. The striatum largely consists of a special type of medium-sized neuron that is inhibitory (that is, it suppresses the activity of connected neurons) and that features many tiny protrusions called spines that receive synaptic inputs from other neurons. When activated, these medium spiny neurons (MSNs) release the inhibitory neurotransmitter molecule GABA (γ-aminobutyric acid) to reduce the activity of other neurons, and together the MSNs control behavioural movement14.
Because MSNs are intermingled with astrocytes and form close contacts with them15, Nagai et al. set out to examine whether MSN activation might affect the activity of surrounding astrocytes. The authors monitored astrocyte activity by making these cells express a genetically encoded calcium indicator — a protein that fluoresces in response to increases in the concentration of calcium ions (which are involved in cell signalling). They found that, when they stimulated MSNs using an electric current, the calcium-ion signalling in nearby astrocytes increased. This increase depended on the release of GABA from the MSNs, and on the activation of type B GABA receptors (GABAB receptors), which are located in the cell membrane and, when bound to GABA, inhibit activity in the rest of the cell. These receptors are examples of a type of membrane protein known as an inhibitory G protein-coupled receptor (GiPCR), which suppresses cell activity by releasing an inhibitory G protein (Gi) inside the cell.
GABAB receptors are expressed by both neurons and astrocytes. The authors set out to dissociate the effects of astrocytic and neuronal GABAB receptors in the striatum, but were unable to deplete these receptors specifically from striatal astrocytes. So, instead, they turned to a tool that mimics the activation of these receptors in astrocytes. They expressed hM4Di, an engineered version of another GiPCR (the human M4 muscarinic receptor), selectively in striatal astrocytes of mice. The hM4Di receptor is selectively activated by a drug called clozapine N-oxide (CNO). Thus, treating these mice with CNO robustly increased calcium-ion levels in hM4Di-expressing astrocytes. The authors observed that CNO treatment of the mice also induced inattentiveness and behavioural hyperactivity — assessed by measuring the animals’ movements and the time spent investigating novel objects, among other behaviours — reminiscent of human ADHD.
Next, Nagai et al. asked how the activation of astrocytic GiPCRs drives behavioural hyperactivity. By examining the firing patterns of MSNs adjacent to astrocytes that expressed hM4Di, they found that CNO treatment boosted the electrical impulses of MSNs and enhanced MSN responses to inputs from neurons in the brain’s cerebral cortex. To unravel the molecular mechanisms that underlie these astrocyte-driven changes in MSN activity, Nagai et al. analysed the levels of RNA transcript molecules in striatal astrocytes, and found that the expression of a molecule called thrombospondin 1 was substantially upregulated in GiPCR-activated astrocytes.
Thrombospondin 1 promotes the formation of new synapses during brain development16. Nagai et al. found that, in the striatum of adult mice, astrocytes can hijack the same molecular mechanism to promote the growth of MSN synapses and thus increase MSN firing. Crucially, blocking thrombospondin 1 signalling — using a molecular inhibitor of the neuronal thrombospondin 1 receptor — prevented CNO-induced increases in MSN synaptic growth and MSN firing as well as CNO-induced behavioural hyperactivity. Collectively, these results suggest that overactivation of striatal astrocytes in adult mice can reactivate a developmental mechanism whereby thrombospondin 1 promotes synaptic growth, resulting in abnormal striatal activity and behavioural hyperactivity.
Nagai and colleagues demonstrate how the acute, specific activation of astrocytes in a particular brain region can lead to ADHD-like behaviour. They illuminate a bidirectional interaction between neurons and astrocytes, through which the two cell types augment each other’s activity (Fig. 1). The fact that this seems to be a positive-feedback-like mechanism might explain why activation of astrocytic GiPCRs can induce behavioural abnormalities so quickly (within 2 hours). The work adds nicely to the growing body of research that demonstrates the importance of astrocytes in brain function and psychiatric disorders5–7.
The study also raises many questions. For example, given that astrocytes are diversified in different brain regions15, is the same GiPCR-induced activation mechanism shared by astrocytes throughout the brain? In addition, the Gs protein-coupled group of receptors, which stimulate cell activity, also boost calcium levels in striatal astrocytes15; are there differences in the spatial and temporal dynamics of the astrocytic calcium signals that are activated by these two seemingly opposite pathways17? If so, do these pathways engage different downstream signalling events and drive different functions?
The striatum contains two subtypes of MSN — one that expresses the dopamine 1 receptor (D1 MSNs) and one that expresses the dopamine 2 receptor (D2 MSNs) — and these subtypes act in opposing pathways to coordinate voluntary movement13. Although Nagai et al. showed that both D1 and D2 MSNs signal to astrocytic GABAB receptors, does thrombospondin 1 from activated astrocytes selectively act on one of these two MSN subtypes to drive movement? These questions await future investigation.
With the advent of modern genetic tools, such as those that enable precise measurement of gene expression in single cells or that allow specific manipulation of different cell populations, future studies will continue to uncover diverse and exciting functions of these star-like cells. Such functions might provide the basis for strategies to treat ADHD and other psychiatric disorders.
Nature 571, 43-44 (2019)