HMN 2026: How groups of neurons support the formation of memories

Neuroscientists and psychologists have been trying to understand how the human brain supports learning and the encoding of memories for over a century. Past studies suggest that memories are stored by groups of brain cells (i.e., neurons), which become active together and produce what is known as an engram.

Engrams are essentially permanent physical and/or chemical changes in the brain associated with the assimilation of specific information or with the formation of new memory associations. A brain region that is known to play a key role in the learning of links between stimuli and outcomes is the CA1 area in the hippocampus.

Researchers at PSL Research University, St Jude’s Children’s Research Hospital and Princeton University carried out a study involving mice that was aimed at better understanding how groups of CA1 neurons contribute to the formation of memories. Their findings, published in Nature Neuroscience, suggest that different groups of cells are active at different phases of learning and encode distinct aspects of experiences.

“My recent paper was inspired by a longstanding question in memory research,” Gisella Vetere, senior author of the paper, told Medical Xpress.

“Many papers had shown the existence of engrams—the specific neurons that store memories—but they often treated memory like a static snapshot, unable to capture the dynamic nature of real experiences. In reality, when we live through an experience, countless neural events occur, but our brain acts like a filter, selecting only the most relevant information.”

Tracking mice neurons during learning

Vetere held some regular meetings with her students at PSL Research University, during which they discussed recent scientific papers. At one of these meetings, they discussed a new experimental technique developed by Christina Kim, who now leads her own lab at Princeton, while she was part of Alice Ting’s lab at Stanford University.

This technique, called FLiCRE, allows researchers to label neurons in the mouse brain that are active at a specific moment in time. Vetere and her studies set out to use this method to determine what cells in the CA1 region were active at different stages of learning and memory formation.

“We immediately realized that this method could act as such a filter: it would allow us to target the most important elements of an experience and reactivate them at will using optogenetics,” said Vetere. “The primary objective of our study was to apply this approach to uncover how ensembles of neurons capture and encode the key aspects of experiences.”

The researchers carried out a series of experiments involving mice that were trained to connect a stimulus with an unpleasant outcome, a process known as associative fear learning.

Using the methods developed by Kim and her collaborators, they closely monitored the activity of neurons in the upper part of the CA1 region during different phases of the associative learning process.

“To investigate the identities of the cells, we used a technique called calcium imaging, which lets us visualize active neurons in real time in freely moving animals,” explained Vetere.

“This allowed us to correlate neuronal activity with specific behaviors or events during a task, identifying cells that are preferentially active at key moments. We then used a complementary technique, FLiCRE, to optogenetically reactivate or inhibit these populations, helping us understand their role in memory recall.”

Vetere and her colleagues observed that different groups of CA1 neurons became active at different stages of the associative learning process. These sets of neurons appeared to encode entirely separate parts of the animals’ experiences. Notably, only some groups of neurons formed engrams and appeared to trigger the activation of fearful memories.

“I think that one of the most remarkable—and somewhat unexpected—findings of our paper is that, during an experience, only a specific subset of neurons is selected to form the memory,” said Vetere.

“Other neurons, even if artificially activated, are not sufficient to elicit the memory, suggesting that their contribution is not critical for the formation of the engram. This highlights how memory formation is a highly selective process, with only certain cells encoding the key elements of an experience.”

Deepening the understanding of memory and psychiatric disorders

The findings of this study offer valuable insight into the neural processes that support the formation of fearful memories in mice, particularly associations between neural stimuli (e.g., sounds) and adverse outcomes (e.g., small electric shocks). If validated in studies involving other mammals and ultimately humans, they could improve the current understanding of some psychiatric and memory disorders.

For instance, anxiety disorders and post-traumatic stress disorder (PTSD) are known to be linked with the formation of unhelpful associations between specific stimuli and traumatic experiences. These learned associations can be highly debilitating for affected individuals, as they can sometimes trigger highly emotional responses, panic attacks, flashbacks or uncomfortable sensations in specific situations.

The recent work by Vetere and her colleagues could help to better understand the processes via which these unhelpful associations are formed in individuals diagnosed with PTSD, anxiety or with other neuropsychiatric disorders. This might eventually guide the development of alternative treatments for these conditions that could contribute to the extinction of fear associations.

“For future research, we aim to explore how these ensembles are reshaped over time and by different experiences, and whether the selected neurons have special molecular features or unique connections that make them key players in memory formation,” added Vetere.

The content is provided for information purposes only.