Astrocytes, Not Neurons, Found to Be Key to Long-Term Memory Stabilization
A groundbreaking new study reveals that brain cells long considered merely supportive players—astrocytes—are, in fact, critical for stabilizing long-term memories. Published on October 15 in the journal Nature, the research offers a surprising new understanding of how the brain retains emotionally charged experiences and could pave the way for novel treatments for conditions like PTSD.
For decades, the prevailing scientific view held that neurons were solely responsible for storing and solidifying memories. However, researchers at the RIKEN Center for Brain Science in Japan discovered that astrocytes, a type of glial cell, play a far more active role than previously thought. These cells appear to “tag” emotionally intense experiences, ensuring they are remembered for days, weeks, or even longer.
The Supporting Cast Takes Center Stage
Astrocytes have traditionally been understood as providing structural support and maintaining the chemical environment for neurons. But when scientists realized that engrams—the physical traces of memories within neurons—couldn’t fully explain the persistence of long-term memories, they began to investigate the potential contributions of astrocytes.
The research team developed an innovative system to monitor astrocyte activity across the entire brain. This system allowed them to fluorescently label astrocytes that were actively producing a protein called Fos, a marker of cellular activation, without detecting signals from neurons. Crucially, the system’s activation was controlled by administering a compound called 4-OHT, enabling precise tracking of astrocyte responses during learning and recall.
How Astrocytes “Tag” Emotional Memories
Using this new technology, the researchers taught mice to associate a specific cage with an unpleasant experience. They observed a key difference between neurons and astrocytes: while neurons showed Fos activity during the initial learning process, astrocytes exhibited strong Fos activity only during the recall of the memory.
Further investigation revealed that this delayed astrocyte activation requires input from neurons in the amygdala—the brain region responsible for processing fear—as well as neurons utilizing noradrenaline, a neurotransmitter involved in the fight-or-flight response. But why the delay?
Single-cell RNA sequencing revealed the answer. In the days following a frightening experience, astrocytes began producing more alpha and beta adrenoreceptors, which are activated by noradrenaline. These receptors act as a “tag,” identifying which astrocytes should be activated when the memory is recalled, effectively reinforcing the memory trace.
Blocking and Boosting Memory Recall
To confirm their findings, the researchers manipulated astrocyte activity. Blocking astrocyte signaling during recall resulted in unstable memories, with the mice showing no recognition of the previously learned association. Conversely, artificially activating astrocytes caused the mice to exaggerate the unpleasantness of even mildly negative experiences and generalize their fear to unrelated environments.
“These findings could lead to new therapeutic approaches that target the astrocytic memory switch, leading to therapies that gently dampen traumatic memories while sparing others,” one researcher explained.
Implications for PTSD and Artificial Intelligence
The implications of this research extend beyond basic neuroscience. The discovery offers a potential new avenue for understanding and treating conditions like PTSD, where emotionally charged memories are often abnormally persistent and easily triggered.
Furthermore, the efficient memory filtering process observed in astrocytes could inspire advancements in artificial intelligence. “Current AI systems are data-hungry and energy-intensive; by learning from astrocytes—which select memories efficiently based on emotional salience and recurrence—we may design more energy-efficient, context-aware AI systems that remember just enough,” the researcher added.
The team’s next steps involve investigating how astrocytes become “eligible” to stabilize memories and exploring the possibility of selectively suppressing or enhancing specific types of memories through astrocyte manipulation.
Source: Dewa, K., et al. (2025). The astrocytic ensemble acts as a multiday trace to stabilize memory. Nature. doi.org/10.1038/s41586-025-09619-2.
