Turning Memory On and Off via Epigenetics

What if a single gene’s epigenetic tag could shift how well you remember something?

In a study led at the Swiss Federal Technology Institute of Lausanne, researchers used CRISPR-based epigenetic tools in mice to dial a gene, specifically Arc, up or down within memory-holding neurons. They found that they could weaken, strengthen or even reverse a memory by editing its epigenetic state.

How memory and epigenetics intersect

Memories don’t live everywhere in the brain. They’re stored in small clusters of neurons called engrams that fire again when a memory is recalled. Research over the past decade has mapped these memory-holding cells and shown how they form lasting links between events, places and emotions. Yet the question of what keeps those connections stable over time remains unknown.

Research has shown that learning leaves epigenetic marks on our DNA – chemical tags which can turn genes on or off. These changes don’t alter the underlying DNA code but can influence how neurons behave. Many studies have linked these broad epigenetic effects to learning and memory, showing that they shape how information is stored in the brain. However, most of that work has focused on large groups of cells or used pharmaceutical modulation to influence many genes at once. That makes it hard to tell if tweaking a single gene can actually change a memory.

“Over the past two decades, several studies have indicated that epigenetic mechanisms – mainly histone acetylation and DNA methylation – may contribute to memory formation, storage and change,” said the authors of the latest study.

The team wanted to go further and find out if switching the epigenetic state of just one memory-related gene, called Arc, inside engram neurons could directly control how a memory is formed, recalled or even erased.

Epigenetic editing of Arc controls memory

The researchers used mice in which memory-encoding neurons could be tagged and manipulated. They combined this system with CRISPR-based epigenetic editing tools, allowing them to make precise adjustments to the activity of Arc inside the neurons. Arc is well known for helping neurons change their connections during learning – a process called synaptic plasticity.

Using harmless viral vectors, the team delivered different versions of the CRISPR system to the hippocampus, a brain region central to memory. One version switched Arc off, the other boosted its activity. The mice then learned to associate a specific context with a mild foot shock, and the team measured how strongly the animals remembered it by tracking freezing behavior.

When Arc was silenced, the animals’ response suggested they barely remembered the shock. When it was activated, their memory strengthened.

Editing also worked on memories that were several days old – normally considered consolidated and resistant to change.

On a molecular level, these changes matched shifts in chromatin – the DNA packaging material – confirming that the team had altered gene activity epigenetically.

Using an anti-CRISPR “off switch,” they were able to reverse the effect within the same animal.

“Such effects occurred irrespective of the memory phase and were reversible within the subject, testifying to their inherent plasticity,” the authors said.

The team’s epigenetic editors were temporally controlled, allowing the memory switch to be applied or reversed at precise time points rather than permanently.

Implications of epigenetic memory control

This work is the first to show that changing the epigenetic state of a single gene can directly control learned behavior. It links the chemistry of gene regulation to the expression of memory, suggesting that the way DNA is packaged inside neurons can act as a kind of “molecular record” of experience.

Advancing Neurodegenerative Disease Research With iPSCs

This eBook explores the latest advances in iPSC-derived models, highlighting how they bridge preclinical and clinical research to accelerate discovery in neurodegenerative disease research.
View eBook

Advertisement

The results point to new ways of studying memory, especially what happens when it breaks down. Findings could help researchers better understand conditions such as post-traumatic stress disorder, addiction and neurodegenerative diseases where memory processing goes awry.

“Similar experimental approaches may in future studies prove useful to better understand other phenotypes... such as in traumatic memories or neurodegeneration, and beyond the field of memory research, where epigenetic mnemonics have been proposed for drug-related memories, as well as for stress-associated behaviors following childhood trauma,” said the authors.

However, the work focused on one gene, one brain region, one type of memory and was only conducted in male mice. How this translates to humans is still uncertain.

Off-target editing and unintended alterations elsewhere in the genome remain a risk to be monitored in future studies.

Next steps will likely test other genes and cell types and explore how many epigenetic sites work together to shape complex memories.

Reference: Coda DM, Watt L, Glauser L, et al. Cell-type- and locus-specific epigenetic editing of memory expression. Nat Gen. 2025. doi: 10.1038/s41588-025-02368-y

This article is a rework of a press release issued by Ecole Polytechnique Fédérale de Lausanne. Material has been edited for length and content.