Astrocytes Are Part of the Circuit: Signaling, Plasticity, and Memory

TL;DR

Astrocytes help maintain the conditions neurons need, but that is only the beginning of the story. They also shape synapses, organize into selective networks, respond to experience, and—in several mouse studies—influence whether a memory can be recalled or maintained. The research does not move memory out of neurons. It gives astrocytes a much more active place in the circuit.

Astrocytes have spent decades being introduced as the brain’s support cells. Fair enough. They clear neurotransmitters, balance ions, help feed neurons, and keep the local environment stable.

But calling them support cells now feels a little like calling a lab manager “helpful.” Accurate—and wildly incomplete.

Astrocyte-neuron signaling is one of the themes in the Neuroscience 2026 program, alongside plasticity, memory, and circuit function. The timing makes sense. Recent mouse studies are showing organized astrocyte networks, learning-associated astrocyte populations, and measurable effects on memory recall and stability. The familiar neuron-centered circuit diagram needs more room.

Why astrocytes have outgrown the supporting role

A synapse may look like a two-cell conversation. Up close, an astrocyte is often standing within earshot.

Fine astrocytic processes extend into synaptic neighborhoods, where they can clear glutamate, buffer potassium, help regulate metabolic support, and respond to local activity. This relationship is often described as the tripartite synapse: the presynaptic neuron, the postsynaptic neuron, and nearby astrocytic processes.

The term is useful, with one caution. It does not mean every synapse uses the same astrocyte mechanism. It means synaptic function depends on the local relationship among all three components. Astrocytes can influence the chemical conditions around a synapse, the availability of fuel, receptor activity, and the molecules that help connections form.

Astrocytes help decide which connections get built

Some of the clearest evidence comes from development. In one primary study of excitatory synapse formation, researchers identified two astrocyte-secreted proteins with opposing effects. One secreted protein, Hevin, promoted synapse formation, while the other protein, SPARC, limited it. Astrocytes were helping set the terms for which neuronal connections formed and matured.

More recent work has widened the view from one synapse to networks that cross brain regions. A 2026 Nature study used a vector-based tracer and whole-brain clearing to follow molecular movement through gap junction-coupled astrocytes in awake, behaving adult male mice. The researchers found selective local and long-range networks. After sensory deprivation, part of the network reorganized.

That result is interesting because the network was neither diffuse nor fixed. It was organized, and it changed with experience. The study did not establish what information traveled through the network, or whether the same pattern appears in females, other ages, or humans. Those questions are still open.

Some astrocytes remember being there

Memory research usually starts with the neuronal engram: a group of connections between cells recruited when a memory forms and reactivated when that memory is recalled. Recent experiments have made the guest list longer.

In a 2025 Nature study, researchers used activity-dependent labeling to identify a subset of hippocampal astrocytes recruited after learning in mice. These astrocytes sat close to engram neurons. When the team experimentally reactivated the learning-associated astrocytes alone, memory recall increased. When they deleted the transcription factor nuclear factor I-A from that astrocyte population, recall was suppressed.

The result does not relocate memory from neurons to astrocytes. It shows that a selected astrocyte population made a causal contribution to recall in this mouse model. That is a more interesting role than background support, and a more precise claim than saying that astrocytes “store memories.”

Memory may leave a multiday astrocyte trace

Another 2025 study followed astrocyte activity across fear learning and recall. An initial experience primed a subset of astrocytes: when the memory was recalled later, those cells responded more strongly. Manipulating their signaling changed the stability and precision of the recalled memory.

The proposed mechanism connects local neuronal engram activity with noradrenergic signaling over time. That longer window matters. Neurons can communicate in milliseconds, while astrocytes may help shape the cellular conditions that persist after an experience and influence what happens when the memory is revisited.

These experiments used mice and specific fear-conditioning tasks. They cannot tell us whether the same mechanism supports a person’s autobiographical memory. What they offer is a testable model: neuronal and astrocytic ensembles may work together across different timescales to form, recall, and stabilize memory.

Keep the claims where the evidence can carry them

Astrocytes have earned a larger place in the circuit diagram. They have not made the neuron obsolete.

They are also far from uniform. Astrocytes differ across brain regions, developmental stages, physiological states, and disease conditions. A hippocampal astrocyte responding during fear conditioning may behave very differently from an astrocyte in the motor cortex during skill learning.

Measurement adds another layer. Calcium signals are widely used to track astrocyte activity, but a calcium event does not automatically reveal what the cell is doing to nearby neurons. Cell specificity, timing, and the strength of a genetic or chemogenetic manipulation can all change the interpretation.

The useful next step is not a broader claim. It is a better-resolved experiment: follow identified neurons and astrocytes across learning and recall, connect structure to function, and test the findings across tasks, ages, sexes, and disease models.

What this means for in vivo neuroscience

Astrocytes make a strong case for studying the brain in context. Combining brain-wide labeling, tissue clearing, single-cell analysis, and animal behavior can reveal organization that any one single preparation may miss. Kent previously covered mouse brain mapping and naturalistic behavior tools; astrocyte research adds another cellular layer to those maps.

Slice electrophysiology still gives researchers excellent control for mechanistic work. In vivo and whole-brain methods answer a different set of questions, especially when long-range relationships or behavioral state matter. Used together, the approaches can connect a molecular mechanism to its place in a functioning circuit.

Good brain data still depends on the animal

The more sensitive the readout, the more the surrounding physiology matters. During cranial-window surgery, stereotaxic injection, implanted recording, or imaging under anesthesia, changes in body temperature, oxygenation, ventilation, or anesthetic depth can become part of the data.

Good physiological control cannot prove the biology, but it can remove avoidable uncertainty about the conditions under which the biology was measured.

If you are planning an in vivo neuroscience workflow, start with the procedure rather than the equipment list. Map the animal’s path from induction through recovery. Decide which variables need to be controlled, which need to be recorded, and where transitions could introduce drift.

Kent’s guide to choosing a rodent anesthesia system walks through the needs of neuroscience procedures, imaging, and longer surgeries. PhysioSuite can combine temperature control with options for pulse oximetry, heart rate, respiration, and end-tidal CO₂. SomnoSuite provides low-flow inhalation anesthesia for mice and rats and can integrate with stereotaxic masks.

Need help matching anesthesia, warming, ventilation, and monitoring to your procedure? Kent Scientific can help you compare options for the workflow you actually run. Final protocol choices should follow the study design, official product instructions, veterinary guidance, and institutional approval.