Forming a memory and keeping that memory stored are very different things -- and they're different at the cellular level in the brain, as well.
Recently, researchers have made progress in understanding how memories are formed, but it is still unclear what controls how well and how long they are stored.
Now, Susumu Tonegawa and colleagues at MIT's Howard Hughes Medical Institute Center for Learning and Memory in Cambridge cast light on this question in today's issue of the journal Cell.
The team created a strain of transgenic mice in which a mutant form of a calcium-dependent enzyme called CaMKIV (Calcium-calmodulin dependent protein kinase) was produced in the forebrain, blocking the normal function of that enzyme specifically in that location.
The transgenic mice seemed normal, but their long-term memories were defective.
The researchers tested both groups of mice in a water maze to see if they could find a hidden platform. The transgenic mice performed as well as the wild type mice during the first couple of days in the maze but, unlike normal mice, they stopped getting better at finding the platform after the second day of training.
As a result, after 12 days, the transgenic mice were significantly worse at finding the platform than their control counterparts.
Other experiments confirmed the transgenic mice were different from normal mice in that although they could form memories, they could not store them for long periods of time.
The team found that the brains of the transgenic mice were unusual in that a protein called CREB (cyclic-AMP response element binding protein), which normally plays a role in regulating gene expression, was not activated.
This suggests that the inability of the transgenic mice to store long-term memories may have to do with the failure to turn on or off specific genes in response to training. This is an important step in understanding how long-term memory works.
[Contact: Susumu Tonegawa]