Random signals in support cells help cement long-term memories

A new study challenges the traditional view that neurons are the sole architects of memory in the brain. Researchers have discovered that astrocytes, a type of star-shaped support cell, generate random electrical signals that are essential for cementing long-term memories. This research, published in the Proceedings of the National Academy of Sciences, suggests that the brain incorporates an element of unpredictability to stabilize neural circuits.

For decades, neuroscientists viewed astrocytes primarily as the “glue” of the nervous system. These glial cells were thought to provide structural support and nutrients to neurons. Over time, this perspective shifted as evidence emerged that astrocytes actively participate in brain signaling. They respond to chemical messengers released by neurons with their own internal calcium flares.

However, astrocytes also exhibit spontaneous activity that does not seem to be triggered by any specific neural input. These calcium fluctuations occur in tiny, localized regions of the cell called microdomains. Because these events appear random, or stochastic, their function has remained a mystery.

A team of researchers led by Gabriele Losi and Beatrice Vignoli, alongside senior authors Giorgio Carmignoto and Marco Canossa, sought to understand if this background noise serves a purpose. They focused on the perirhinal cortex. This region of the brain is responsible for recognition memory, such as identifying familiar objects.

The investigators hypothesized that these random signals might influence how the brain consolidates memories over time. Memory consolidation is the process by which a temporary memory trace is stabilized into a long-lasting one. This often depends on a phenomenon called long-term potentiation.

Long-term potentiation refers to the persistent strengthening of synapses, the connections between neurons. When neurons fire together repeatedly, their connection becomes stronger. This synaptic strengthening is the cellular foundation of learning.

To test their hypothesis, the team used advanced imaging techniques to observe astrocyte activity in mouse brain tissue. They utilized a calcium indicator that glows when calcium levels rise inside the cell. This allowed them to track the flickering activity of the microdomains in real time.

The researchers confirmed that these calcium flashes occurred spontaneously. They persisted even when the researchers used toxins to silence the electrical firing of nearby neurons. This proved that the astrocytes were generating these signals independently.

Next, the team investigated whether this spontaneous activity influenced synaptic strengthening. They stimulated the neurons to induce long-term potentiation. Under normal conditions, the synaptic connection remained strong for hours.

The researchers then introduced a genetic tool to dampen the calcium signaling within the astrocytes. When the spontaneous calcium flashes were suppressed, the synaptic strengthening collapsed. The connection initially grew stronger but faded back to baseline levels within roughly an hour.

This result implied that while neurons can initiate a memory trace, they cannot maintain it without the help of astrocytes. The researchers pinpointed the molecular mechanism behind this failure. They looked at a protein called brain-derived neurotrophic factor, or BDNF.

BDNF acts like a fertilizer for brain cells, promoting growth and survival. The study revealed that the spontaneous calcium flashes trigger the astrocytes to release BDNF. This protein then binds to specific receptors on the neurons called TrkB receptors.

Sustained activation of these receptors is required to lock in the changes at the synapse. The random, recurring nature of the astrocyte signals ensures that BDNF is released over a prolonged period. This extends the window of time for the neurons to solidify their connection.

To prove that BDNF was the missing link, the scientists applied the protein directly to the brain tissue where astrocyte activity was blocked. The external supply of BDNF rescued the synaptic strengthening. The memory trace persisted just as it would in a healthy brain.

The team then moved from tissue samples to live animal behavior. They used a standard memory assessment known as the object recognition test. In this task, mice spend time exploring two identical objects.

Later, one of the objects is replaced with a new one. Mice with normal memory will spend more time exploring the novel object. This indicates they remember the familiar one.

The researchers engineered the astrocytes in the mice to allow for temporal control. They could switch off the spontaneous calcium signals using a specific chemical drug. This allowed them to disrupt astrocyte activity at precise moments during the memory process.

When the researchers inhibited the astrocytes immediately after the mice learned the objects, the animals failed the test 24 hours later. They explored both objects equally, indicating they had forgotten which one was familiar. The initial learning had occurred, but the long-term memory had not formed.

However, if the researchers waited to inhibit the astrocytes until hours after the learning event, the memory remained intact. This demonstrated that the spontaneous activity is required only during a critical window following the experience. The astrocytes provide the necessary chemical support to stabilize the circuit while the memory is fresh.

The study highlights the stochastic nature of this process. The calcium events in the microdomains do not follow a set pattern. They are inherently unpredictable.

The authors propose that this randomness is not a flaw but a feature. It may introduce a probabilistic element to memory storage. By randomly engaging different parts of the astrocyte, the brain might select which synaptic connections are worth preserving.

This mechanism ensures that not every fleeting neural activation becomes a permanent memory. Only those connections that receive the sustained chemical support from astrocytes will endure. The random activity acts as a filter for information retention.

The findings also clarify the relationship between the evoked responses and spontaneous ones. When neurons fire rapidly, they can trigger a large calcium response in the astrocyte soma, or main body. But this main response is not enough for consolidation.

The localized, random microdomain flashes are distinct from the global cell response. They operate autonomously. This adds a layer of complexity to how non-neuronal cells process information.

There are caveats to consider in this research. The study was conducted in mice, and human brain physiology differs in complexity. Whether this specific mechanism operates identically in humans remains to be verified.

Additionally, the exact source of the randomness requires further exploration. While the signals appear stochastic, underlying intracellular processes likely govern their frequency and distribution. Understanding these drivers is a necessary next step.

Future research will likely focus on how this mechanism applies to other types of memory. The perirhinal cortex handles object recognition, but other areas like the hippocampus manage spatial and episodic memory. Astrocytes in those regions may behave differently.

The researchers also aim to investigate the implications for brain disorders. Issues with memory consolidation are hallmarks of conditions such as Alzheimer’s disease. Dysfunctional astrocyte signaling could be a contributing factor.

If the spontaneous activity of astrocytes is dampened in neurodegenerative diseases, it could explain why new memories fail to stick. Restoring this signaling could theoretically offer a therapeutic pathway. This is a speculative but promising avenue for future medical research.

Ultimately, this work elevates the status of the astrocyte. It suggests that our ability to remember the past depends on the random flickering of cells once thought to be passive bystanders. The brain’s stability appears to rely on a fundamental element of chaos.

The study, “Spontaneous activity of astrocytes is a stochastic functional signal for memory consolidation,” was authored by Gabriele Losi, Beatrice Vignoli, Rocco Granata, Annamaria Lia, Micaela Zonta, Gabriele Sansevero, Francesca Pischedda, Angela Chiavegato, Spartaco Santi, Lorena Zentilin, Nicoletta Berardi, Gian Michele Ratto, Giorgio Carmignoto, and Marco Canossa.