Frustrated Fish Reveal Glia’s Computational Role in Motivated Behavior

New research shows that a population of astrocytes in fish keeps track of failed attempts to swim and signals when to give up.

The majority of cells in your brain aren’t neurons; they’re glia. According to the traditional view, the duties of these cells are mainly “supportive”: they control energy levels, fight disease and clean up excess neurotransmitters while neurons perform the important computations for behavior. A new study in zebrafish, however, found that glial cells play a crucial computational role in motivated behavior.

When a zebrafish swims forward, it expects to see the visual world around it move backward as a result. In the new study, published in Cell in June, Misha Ahrens, a neuroscientist at Janelia Research Campus and investigator with the Simons Collaboration on the Global Brain, and colleagues paralyzed the muscles of larval zebrafish so that they could no longer move forward. After several seconds of futile attempts, the fish stop trying. They remain in this passive state for several seconds before attempting to swim again.

Calcium activity in radial astrocytes peaks after a futile swim attempt (marked by white lines at bottom.) Mu et al. Cell 2019.

To uncover the cellular mechanisms behind this pattern of behavior, the team performed whole-brain imaging of calcium levels in both neurons and glia. Through this imaging, they identified a group of glial cells known as radial astrocytes that were partially located in the lateral medulla oblongata. The signal from these cells correlated with the switch from swimming to passivity. Specifically, the amount of calcium in these cells increased as the fish made failed attempts at swimming and peaked just after the fish stopped trying.

To test if these cells play a causal role in the decision to give up, the researchers killed a portion of them and checked for changes in behavior. Fish with these cells missing spent half as much time in the passive state as intact fish. Artificially activating these astrocytes, either chemically or optogenetically, increased passivity, even in circumstances in which the fish’s swimming was effective in eliciting the expected visual feedback.

The researchers next aimed to find out how the astrocytes communicate with neurons to induce this behavior. They found that artificially activating the astrocytes triggered nearby inhibitory neurons, providing a potential means by which the astrocytes could control behavior. Indeed, the researchers found that activating these inhibitory neurons directly suppressed swimming. How exactly the astrocytes activate these neurons remains to be determined, but astrocytes are known to secrete signaling molecules detectable by neurons.

The next step was to identify the cause of increased activity of the astrocytes. The researchers identified neurons in the medulla oblongata that were active each time the fish attempted to swim but failed, and this activity preceded calcium increases in astrocytes. These neurons release norepinephrine, a neuromodulator that astrocytes can detect. In fact, the researchers identified the exact receptors on these astrocytes that respond to norepinephrine and cause an increase in calcium.

In total, the researchers demonstrated that the astrocytes use the firing of the norepinephrine neurons to keep track of how many recent swim attempts failed — and integrate this information to suppress further attempts. Sufficiently intense activity in the astrocytes activates the inhibitory neurons, preventing swimming for a short period.

It’s unclear at this point whether these findings will generalize to mammals, but this thorough exploration in zebrafish adds to a growing body of literature indicating that glial cells offer more than just support for the important computations of the brain.