Our mind often seems like a gigantic library, where memories are written on parchment and stored away on shelves. Once filed, they remain steadfast and inviolate over time, although some may eventually become dusty and forgotten.
Now, Reut Shema, Yadin Dudai and colleagues from the Weizmann Institute of Science have found evidence that challenges this analogy. According to their work, our memory is more like a dynamic machine - it requires a constant energy supply to work. Cut the power and memories are lost.
Shema found that the plug that powers our memories is an enzyme called PKMzeta. This molecule is vital for a process called long-term potentiation (LTP) where a the strength of a synapse - the connection between two nerve cells - is increased in the face of new experiences. This process, and thus PKMzeta, fuels the production and storage of new memories.
Shema demonstrated the importance of PKMzeta by inactivating it in the brains of rats. He trained the rats to avoid the taste of the artificial sweetener saccharin and then injected the part of their brains that control taste with a chemical called ZIP that stops PKMZeta from doing its thing.
The results were striking - ZIP erased the rodents' memories of what they'd learnt. It even killed off the relevant memories when the rats were injected a month after their training. In human terms, that's the equivalent of erasing memories that were several years old.
Even if the rats were given trials to reinforce their aversion to saccharin, they forgot all about it once ZIP was brought into play. Rehearsal didn't 'immunise' them against the loss of PKMzeta.
More surprising still, the process seemed to be irreversible, at least within the duration of the experiment. Twelve days after ZIP was administered, the rats still had no recollection of the supposed unpleasant taste of saccharin. Their memories had not just been clouded over, they seemed to have been truly erased.
ZIP has no effect on how well the rats created new memories - if they were injected before they were taught to avoid saccharin, they picked things up just as well. But it seems that PKMZeta is vital for the continuing existence of new memories.
If results can be generalised to other parts of the brain, and indeed, to humans, they suggest that memories are not simply writ once on our mental network and left to be retrieved. They exist because of ongoing processes in the brain. Long after they are created, memories are still incredibly vulnerable to loss, perhaps even irreversibly so.
This impermanence may actually be beneficial - it could render the entire network more flexible and make it easier for new experience to be added to the mix.
Reference: Shema, Sacktor & Dudai. 2007. Rapid erasure of long-term memory associations in cortex by an inhibitor of PKMzeta. Science 317: 951-953.