Illustration: Jim Cooke

Researchers from the University of Oxford have rewritten positive memories associated with cocaine in mice. The achievement could expand our understanding of memory, while demonstrating that it’s possible to neurologically reverse ingrained bad behavior, such as drug addiction.

Neuroscientists Stéphanie Trouche and David Dupret from Oxford’s MRC Brain Network Dynamics Unit trained mice to prefer a particular location using cocaine. Then they altered those positive associations using optogenetics—a genetic technique in which living brain cells can be manipulated or controlled with light (typically via fiber optic cables). The mice lost their preference for the cocaine-associated environment, suggesting their memory had been rewritten. The results of this experiment can now be found in Nature Neuroscience.

Whenever a mammal is in a particular environment, the specialized neurons associated with a particular place in its hippocampus (the part of the brain responsible for long- and short-term memory consolidation and spatial navigation) become active. Place cells enable an animal to build a mental representation of a specific location. These cognitive maps can be tied to an emotional response based on an animal’s prior experience at a particular location. For example, if a mouse encounters a hostile cat in a specific area, its cognitive map will feature a negative association, expressed as fear. Alternately, the discovery of a tasty meal would imprint a positive association with that location.


These associations are stored in the brain in the form of memory engrams. Back in 2013, a research team from MIT learned that it’s possible to implant false engrams in a mouse’s mind, causing it to recall an experience that never occurred. After training a mouse to be fearful of a particular room using electric shocks, the researchers activated a light-sensitive protein in its brain using optogenetics.

Credit: Shutterstock

When the mouse was placed in a room known to be safe, the light triggered the memory associated with the dangerous room, causing it to freeze in fear. This suggests it had formed a false-memory associated with an entirely different room. The experiment affirmed the longstanding notion that memories are physically resident in the brain, and that they’re subject to manipulation.


The new Nature Neuroscience study shows that these location-specific memories are also subject to permanent recoding. Expanding on the work done at MIT, Trouche and Dupret performed an experiment to see if it was possible to overwrite an animal’s association with a particular environment as a means to halt maladaptive behavior. Heading into the experiment, the researchers weren’t sure if place cells could be selectively edited in such a precise way, as each can contribute to multiple spatial representations. But they found that this is very much possible.

Using cocaine, the researchers trained mice to prefer a particular environment. When given the choice between hanging out in a cocaine-associated environment or one linked to a bland saline solution, the mice (not surprisingly) preferred to loiter at the cocaine-linked area.

Light-sensitve tagged proteins (top) compared to a control. Credit: MRC Brain Institute

Next, the scientists labelled the place cells in the brains of mice that were active when they were in the cocaine-paired environment, so that they would express light-sensitive proteins. While the mice were exploring the cocaine-paired environment, the researchers switched on the light, which silenced the tagged neurons.

At the same time, this caused a batch of unmasked “quiet neurons,” i.e. unassociated place cells, to become active, which enabled the emergence of an alternative cognitive map. The mice instantly lost their preference for the cocaine-linked location, suggesting that their memory had been rewritten, or recoded, in such a way that the environment was no longer associated with the drug.

This result affirms the engram theory of memory consolidation, suggesting that our memories are stored as biophysical or biochemical changes in the brain in response to external stimuli and experiences. It also demonstrates that certain behaviors, in this case maladaptive behaviors caused by drug addiction, can be manipulated, and even rewritten, using optogenetics. But it’s not entirely clear how applicable this work is to humans, given that optogenetics requires genetic manipulation and the use of invasive fiber optic implants in the brain.

Regardless, these insights could inspire the development of similar interventions delivered by drugs or other means. As the researchers concluded in their study, “[Our] data demonstrate that recoding a selective hippocampal engram represents a potent strategy to reset spatial memories and neutralize maladaptive behavior.”

[Nature Neuroscience]

Email the author at and follow him @dvorsky.