How to build a brain

Study shows it takes complex crosstalk between astrocytes, microglia, and neurons.

Neurons, the brain’s main signal-senders, connect tightly to one another through structures known as synapses. These synapses aren’t floating alone in the brain; instead, they are embedded in a “sea” of star-shaped cells called astrocytes. Astrocytes were once dismissed as mere “glue” and support for neurons, but cell biologist Cagla Eroglu, PhD, has spent her career revealing their surprising powers. Her team’s work has shown that they actively shape how the brain forms and organizes its circuits.

Now, her lab at Duke University School of Medicine has uncovered another layer of complexity: a three‑cell partnership, with astrocytes at the center, that trims away unneeded neural connections in the visual cortex shortly after birth — a pruning process essential for healthy brain development. “Brain development is not only guided by neuron-neuron interactions; it's crosstalk between astrocytes, microglia, and neurons,” she said. 

Published in the journal Neuron, the study points toward a better understanding of sensory processing disorders and neurodevelopmental conditions, including autism.

During early development, the visual cortex forms many synapses, but soon after birth, the brain begins pruning them. Scientists don’t yet fully understand why, but they do know the process helps fine‑tune this busy region of the brain, said Eroglu, the Duke Health Distinguished Professor of Cell Biology.   

Because the visual cortex must blend visual input with data from touch and hearing, disruptions in this fine‑tuning may contribute to conditions marked by heightened or reduced responses to sensory input.

This study investigated a protein called Hevin, which is released by astrocytes. Hevin is already known to help build certain kinds of synapses in the visual part of the mouse brain.

The researchers found that a previously unknown form of Hevin – a version that is cleaved in two – helps prune synapses.

“We thought we had already figured out everything Hevin is doing,” Eroglu said. “But of course that’s not how biology works. Humans want to simplify things, but biology uses all the resources it has to generate the complexity of the brain.”

The Two Faces of Hevin

A former graduate student in the Eroglu Lab, Juan Ramirez, PhD, discovered this new role for Hevin when he became curious about the timing of thalamocortical synapse formation in this region. “We started to see really strange stuff happening,” Eroglu said.

Using high‑resolution microscopy, Ramirez noticed that Hevin was highly expressed at these synapses in mice right around a crucial time window — just as eyes begin to open at 12-14 days old. Hevin-rich synapses stayed; Hevin-poor synapses disappeared.

Suspecting that immune cells called microglia were responsible for removing the synapses, the team used a genetic tool to eliminate microglia from the mouse brain. Without microglia, the synapses could not be eliminated.

To test whether Hevin acted as a “stay” or “go” signal, the researchers then isolated microglia in a culture system and treated them with purified Hevin. Instead of calming the cells, Hevin activated them: in isolation, it pushed microglia to go into an “eat and digest” mode. That was the opposite of what they saw in the intact brain.

That result led the team to propose that Hevin must exist in at least two forms — one that stabilizes synapses, and another that signals their removal.

"Humans want to simplify things, but biology uses all the resources it has to generate the complexity of the brain.” — Cagla Eroglu, PhD

Then they found that, indeed, the cleaved version of Hevin activates microglia by binding to a receptor on their surface — TLR4. Once activated via TLR4, microglia shift into a specialized state that enhances their ability to break down and remove synapses. This signaling is essential for normal synapse pruning in young mice, the team found.

Eroglu credits a collaboration with Duke psychologist and neuroscientist Staci Bilbo, PhD, with helping her team study microglia in depth.

This discovery suggests that studying Hevin is more important than ever. “With Hevin performing two functions, you are controlling the synaptogenesis and elimination through the supply of the same molecule,” Eroglu said.

Further study could contribute to the understanding of neurodevelopmental disorders. For instance, SPARC-like 1, the gene that makes Hevin, is linked to autism, and aberrant microglial activation in the brain, especially due to prenatal infections in the mother, has also been linked to autism-like disorders.

In addition, Eroglu is excited about investigating whether the specific type of microglia they discovered can contribute to disease states. “We think this dual role of Hevin may be hijacked when the brain is under stress, tipping the balance toward synapse loss,” Eroglu said. “We are pursuing this direction in our ongoing studies of diseases where synapses are lost.”

Funding: the Adelson Medical Research Foundation and the National Institutes of Health.

Share