Our experiences leave traces in the brain, stored in small groups of cells called “engrams”. Engrams are thought to hold the information of a memory and are reactivated when we remember, which makes them very interesting to research on memory and age- or trauma-related memory loss. At the same time, scientists know that the biology of learning is accompanied by epigenetic changes, which refer to the ways the cell regulates genes by adding chemical "post-it notes" on DNA. But the question of whether the epigenetic state of a single gene can, in turn, cause a memory to change has thus far remained unanswered.
In the current study, Davide Coda combined CRISPR-based gene control with a technique that tags engram cells in mice. He focused on Arc, a gene that helps neurons adjust their connections to other neurons. By targeting the control region of Arc, he asked whether flipping its epigenetic “switch” could directly change memory. Coda developed specialized, CRISPR-based tools that could either dial down or boost Arc activity in memory neurons. Some, like the KRAB-MeCP2 tool, were designed to switch off gene activity by adding repressive marks that make the DNA less accessible, while others opened the DNA and turned the gene on. These tools were essentially an “epigenetic switch” for the Arc gene. The HFSP-supported fellow then used harmless viruses to deliver these tools directly into the hippocampus of mice, a brain region central for storing and retrieving memory. The mice were then trained to link a specific place with a mild foot shock. By changing the epigenetic state of Arc in neurons, Coda and his mentor could see whether the animals remembered the shock or not. The scientists also added a “safety switch” that could undo the editing and reset the memory state.
The study showed that epigenetically silencing Arc in engram cells prevented mice from learning, while boosting it made their memory stronger. These changes could be reversed in the same animal, showing that this epigenetic “switch” can dial memory expression up or down. Even memories that were already several days old, which are usually hard to change, could be modified. At the molecular level, the editing caused changes in Arc’s activity and DNA packaging that matched the behavioral effects.
The study is the first direct demonstration that changing the epigenetic state in memory cells is necessary and sufficient to control memory expression. It points to new ways of exploring how memories are stored and altered, which could eventually also be relevant in humans. In the future, similar approaches could help researchers better understand conditions where memory processing goes away, such as traumatic memories in PTSD, drug-related memories in addiction, or the memory problems that appear in neurodegenerative diseases.